optical communication based research papers

• Reference Manager
• Simple TEXT file

People also looked at

Review article, artificial intelligence in optical communications: from machine learning to deep learning.

• State Key Laboratory of Information Photonics and Optical Communications, Beijing University of Posts and Telecommunications, Beijing, China

Techniques from artificial intelligence have been widely applied in optical communication and networks, evolving from early machine learning (ML) to the recent deep learning (DL). This paper focuses on state-of-the-art DL algorithms and aims to highlight the contributions of DL to optical communications. Considering the characteristics of different DL algorithms and data types, we review multiple DL-enabled solutions to optical communication. First, a convolutional neural network (CNN) is used for image recognition and a recurrent neural network (RNN) is applied for sequential data analysis. A variety of functions can be achieved by the corresponding DL algorithms through processing the different image data and sequential data collected from optical communication. A data-driven channel modeling method is also proposed to replace the conventional block-based modeling method and improve the end-to-end learning performance. Additionally, a generative adversarial network (GAN) is introduced for data augmentation to expand the training dataset from rare experimental data. Finally, deep reinforcement learning (DRL) is applied to perform self-configuration and adaptive allocation for optical networks.

Introduction

Machine learning (ML) techniques have been developed and applied to optical communication in both the physical layer and network layer for years ( Musumeci et al., 2018 ; Khan et al., 2019 ). Various algorithms from ML communities powered a wide range of aspects in optical communication, involving digital signal processing (DSP), optical performance monitoring (OPM), signal detection and analysis, proactive fault management, network automation, and optical sensing, etc. The conventional ML system is limited by the ability to undertake feature extraction and complex analysis, and has always relied on considerable domain expertise and feature engineering. In recent years, rapid advances in information technology have made great strides and parallel developments in computation and low-cost computing hardware have made big data modeling possible. Driven by this growth in the volume of data and improvements to computing power, ML has successfully evolved into deep learning (DL), which addresses complex and large-scale problems with robust, adaptable, and efficient solutions ( LeCun et al., 2015 ), as illustrated in Figure 1 .

Figure 1 . Advances in artificial intelligence in optical communications. Driven by powerful parallel computing capacity and big data, traditional machine learning algorithms are progressing to deep learning techniques with a variety of applications, promoting the evolution of optical communications toward intelligence.

In general, DL can be understood as a deep neural network (DNN) with multiple non-linear layers made up of a large number of neurons, each of which is mathematically modeled as an activation function. In DL communities, different algorithms with specific structures were suitable for different problems and specialized in different data types. Among them, convolutional neural network (CNN), recurrent neural network (RNN), generative adversarial network (GAN), deep reinforcement learning (DRL), end-to-end learning based on autoencoder, and their variants have made a distinctive contribution to fields such as machine vision, natural language processing, drug discovery, genomics, speech recognition, information retrieval, affective computing, and automatic deriving ( Deng, 2014 ). Meanwhile, to promote the development of artificial intelligence (AI) in optical communication, the evolution from ML to DL is making major advances in a wide variety of applications in both physical and network layers ( Fan et al., 2020 ; Häger and Pfister, 2020 ; Saif et al., 2020 ).

This paper reports the progress of AI in optical communication from ML to DL. Unlike other review papers about conventional ML algorithms, the presentation focuses on state-of-the-art DL techniques and aims to highlight the contributions of DL to optical communication for both the physical layer and the network layer. Examining the characteristics of different DL algorithms and data types, we briefly review multiple DL-enabled applications for optical communication. First, as one of the most popular DL algorithms, CNN is introduced for image recognition to process seven kinds of common image data from optical communication to execute various functions. Then RNN is applied for sequential data analysis to process digital signal waveforms, network traffic data, and equipment state parameters. In addition, a data-driven channel modeling technique using DL is proposed to provide a supplementary solution to the conventional block-based modeling, which could also improve end-to-end learning performance. As an emerging technique, GAN is implemented for data augmentation to expand image data and network traffic data. Finally, DRL is considered for various decision-making tasks, including routing, resource allocation, and automatic configuration.

Convolutional Neural Network for Image Data

DL belongs to a branch of the ML family mainly referring to the faction of neural networks. The term “neural network” has its origins in attempts to find mathematical representations of information processing in biological systems, which are built of a lot of interconnected neurons. As the basic unit of a neural network, each neuron can be modeled as an activation function to emulate the process of transferring information in the practical biological system. According to the network topology, neural networks can be categorized into feedforward networks and feedback networks. A convolutional neural network is a specialized type of feedforward network for primarily processing image data that can be regarded as a two-dimension (2D) grid of pixels ( LeCun et al., 2015 ). The operating process of CNN can be described as convolution, pooling, and activation.

Convolution

The kernel convolves with pixel points across the width and height of the input image, computing the dot product between the entries of the kernel and input. The kernel works like a filter that scans the input image to extract the informative features for recognition. The extracted features from the image are displayable and explainable, such as eyes, nose, or mouth in face photos. Convolution takes advantage of sparse interaction, parameter sharing, and equivariant representations to improve the performance of image recognition.

Pooling: Down-Sampling Operation

The output of the convolution layer at a certain location is replaced by a summary statistic of the nearby outputs. The typical pooling is to calculate the average or maximum value of a small local region in one feature map to down-sample the dimension of the feature map, thereby greatly reducing the parameter size and creating an invariance to small translations of the input.

Activation: Non-linear Operation

The representation capacity of the whole network is improved through the non-linear mapping between adjacent layers. Common activation functions include ReLU, Softmax, Softplus, and Sigmoid, etc.

Due to the above factors, CNN is particularly effective at examining image data, including image recognition, objection detection, image understanding, and video translation ( Gu et al., 2018 ). It has been statistically established that images often account for a large proportion of various data types. Therefore, CNN is one of the most useful approaches in DL for image processing. In optical communication, most data are denoted in the format of a digital signal, while some other kinds of information are presented in the form of images, as summarized and displayed in Figure 2 . Compared with the data format of digital vectors, one great advantage of image formats is that various digital data of different sizes can be comprehensively and integrally presented in a picture with a fixed pixel size. Image data with a fixed size can therefore contain various information, which is important for ML and DL in keeping their structures stable ( Wang et al., 2019 ).

Figure 2 . Application of convolutional neural network (CNN) in optical communication for image processing. (A) Summarization of image data in optical communication: linear polarization (LP) mode diagrams, orbital angular momentum (OAM) mode diagrams, eye diagrams, constellation diagrams, asynchronous delay-tap plot (ADTP) diagrams, asynchronous amplitude histograms (AAH) diagrams, and optical spectrum diagrams. (B) The structure of CNN is composed of convolution layers, pooling layers, and fully-connected layers. (C) A variety of functions can be achieved by CNN for optical communication.

As can be seen from Figure 2 , the seven kinds of typical image data in optical communication are linear polarization (LP) mode diagrams, orbital angular momentum (OAM) mode diagrams, eye diagrams, constellation diagrams, optical spectrum diagrams, asynchronous amplitude histograms (AAH) diagrams, and asynchronous delay-tap plot (ADTP) diagrams (ADTP combines asynchronous sampling with a two-tap delay line, so that each sample point comprises two measurements, separated by a fixed time corresponding to the delay length. The samples are plotted as sample pairs, producing a joint map of the power and evolution over the delay time) ( Wang et al., 2017a , b ; Li et al., 2018 ). Through analyzing and processing these image data, CNN can explore a large number of informative features for optical communication to execute a variety of functions, including but not limited to channel estimation, mode demodulation, optical signal analysis, impairment diagnosis, OPM, DSP, and spectral analysis. For example, CNN is capable of: detecting mode crosstalk and estimating a few mode fiber channels from LP mode diagrams; demodulating multiplexed modes and detecting atmospheric turbulence from OAM mode diagrams; analyzing the signal quality; diagnosing system impairments from eye diagrams (for intensity-modulated signals) and constellation diagrams (for complex-modulated signals); monitoring optical-to-noise ratio (OSNR) and identifying modulation format with low-cost methods from ADTP and AAH diagrams; and measuring and analyzing spectral characteristics from spectrum diagrams.

Recurrent Neural Network for Sequential Data

Unlike CNN designed for image data, RNNs are specifically proposed for sequential data, where temporal correlations exist at a range of different timescales. Different from feedforward neural networks, RNNs containing cyclic connections aim to provide neural networks with memory, meaning that the outputs are not only related by the current inputs but also the formerly available information ( Mikolov et al., 2010 ). Thus, RNNs have achieved great success in sequence modeling and prediction tasks, such as speech recognition, handwriting recognition, language translation, and stock price forecasting. The principle of RNN is illustrated in Figure 3 . The input vector is a series of sequential data X = {… x t −1 , x t , x t +1 …}, and the neurons in the hidden layer get inputs from not only x t of the input layer but also the output h t −1 of the hidden layer at the previous time steps. Passing through multiple hidden layers, an input sequence x t can be mapped into an output sequence y t that involves some previous stated information.

Figure 3 . Application of recurrent neural network (RNN) in optical communication for sequential data processing. (A) Summarization of sequential data in optical communication: digital signal waveforms, network traffic data, and equipment state operating parameters. (B) The schematic of RNN considers the extracted features in the previous state as the one of the current input information and the current outputs depend on the current and previous inputs to provide the neural network with memory; and Long short-term memory (LSTM) block diagram is the variant of RNN that can learn long-range temporal relationships among sequential data. (C) A variety of functions can be achieved by RNN for optical communication.

However, conventional RNN finds it difficult to learn long-term dependencies from sequential data. To overcome this weakness in RNNs, long short-term memory (LSTM) was designed to learn long-range temporal relationships among sequential data and remember inputs for a long time ( Zia and Zahid, 2019 ). As one of the most famous RNN variants, the core idea of LSTM is the memory cell, which can pass information through time steps, and structures called gates, which are used to remove or add information to the memory cell, as shown in Figure 3B . The operating process of LSTM can be summarized by forgetting the old state and memorizing the fresh state such that the useful information in the cell can be passed on, and the useless information can be discarded. Thus, LSTM can not only allow the accumulation of information over a long period of time but also forgets the old state by setting it to zero and starting to count afresh.

In the era of big data, except for image data, most of the rest are sequential data, such as speech, language, and words. In optical communication, most data are sequential data, such as optical and electrical signals, network traffic data, equipment state operating parameters, as summarized and displayed in Figure 3A . In optical communication, for tasks that involve these sequential data, it is better to use RNNs to realize digital signal pre-distortion and post-compensation, inter-symbol interference (ISI) cancellation, network traffic prediction, and equipment failure management, etc.

The optical signals can be regarded as a series of time-domain data, and the mutual influence and the experienced impairments from the transmission process can be embodied into temporal signal waveforms. Considering the superior performance of RNN for these data, RNN can pre-distort signal before transmission to resist transmitter imperfection and the post-compensate signal after receiver to mitigate system impairments or identify the crosstalk between adjacent symbols to cancel the ISI ( Lu et al., 2019 ; Deligiannidis et al., 2020 ; Zhao et al., 2020 ).

For network traffic data, the traffic loads fluctuate regularly or irregularly over time according to daily statistics ( Lu et al., 2015 ). Based on previous scenes, RNN can build a prediction model for large-scale network traffic forecasting from the perspective of temporal analysis, which is important for load balancing and network planning ( Gui Y. et al., 2020 ; Zheng et al., 2020 ).

Early-warning and proactive protection are becoming increasingly critical for network operators as a failure of the optical network could result in huge economic loss. The operating conditions of network equipment can be reflected in the equipment state parameters, which are varied over time. Through analyzing a great deal of historical data, RNN can learn the variation trend of state parameters and establish a failure prediction mechanism to prevent risk in advance ( Wang et al., 2018 ; Zhang et al., 2020 ).

End-To-End Learning for Joint Optimization With Dl-Based Channel Model

The conventional model of the optical communication system is constructed in a divide-and-conquer manner and consists of a series of model blocks, including symbol mapping, shaping filter, laser, modulator, fiber channel, amplifier, optical filter, detector, low-pass filter, and digital sampling, as shown in Figure 4A . This block-based optical communication system is strongly dependent on practical channel conditions and is characterized by rigid mathematical models ( Agrawal, 2012 ). However, the conventional block-based communication systems still have the following deficiencies: (a) they are only effective in tractable and stable scenarios, but invalid for those complex and dynamic scenarios; (b) they require a lot of artificial expertise; and (c) they have a relatively long computation time owing to the small step sizes and repeated iterative operations they undertake.

Figure 4 . Deep learning for optical communication modeling. (A) The conventional block-based optical communication system, constructed in a divide-and-conquer manner using a series of model blocks. (B) Deep learning-based optical communication model, built by the data-driven multi-layer neural network. (C) Schematic of end-to-end learning for optical communication, based on the DL-based channel model.

In deep learning communities, autoencoder is another important and popular algorithm. It is an unsupervised learning algorithm for a neural network that sets the target output values to equal the inputs. The autoencoder has been applied in dimensionality reduction, feature reconstruction, and data encryption ( Tschannen et al., 2018 ). A new fundamental way to interpret the entire communication systems as an autoencoder has been proposed. It was first presented in wireless communication systems before being introduced to optical communication systems ( Karanov et al., 2018 ). This technique is based on the concept of end-to-end learning that seeks to jointly optimize the transmitter and receiver components in a single process. However, a major drawback hindering practical implementation is that a differentiable channel model is necessary to execute parameter adjustment through backpropagation. Accordingly, a DL-based fiber channel modeling scheme was proposed ( Wang et al., 2020 ). In theory, DL can approximate any function to solve both linear and non-linear problems. According to the characteristics of DL, the model functions can be approximated by mapping independent to dependent variables, corresponding to the input and output data as shown in Figure 4B . DL constructs an approximate model for a black box driven by source data and received data. Furthermore, because the scheme does not rely on expert experience, it can significantly reduce the modeling cost and improve the simulation efficiency. This transmission simulation model in the DT system can not only digitize the physical process but also provide the numerical channel model that is important for adaptive damage compensation, like the end-to-end learning method, to ensure high reliable transmission of optical communication. Based on the idea of an auxiliary channel, a DL-based channel as shown in Figure 4C was also flexibly embedded into an end-to-end learning model to perform joint optimization more accurately ( Karanov et al., 2020 ; Li M. et al., 2020 ).

Generative Adversarial Network for Data Augmentation

One of the main motivations for DL is having an effective and available dataset for training, and more adequate data contribute to a better generalization of the model. However, in practice, labeled data are valuable and rare. In optical communication, it is difficult to collect both image data and sequential data, particularly experimental data and practical data from network operators or corporations. In addition to guaranteeing sufficient data, diversity is also essential to improving the robustness and generalization of DL models. Therefore, a lack of sufficient and diverse training data is one of the major limitations on DL to be well-applied in optical communication.

GAN was recently introduced as an emerging technique to implement data augmentation. At first, GAN was proposed by Ian Goodfellow et al . as a way to generate image data, including handwritten digits, human faces, and animal images ( Goodfellow et al., 2014 ). The idea behind GAN was based on the concept of zero-sum game theory, as shown in Figure 5 . The framework of GAN consists of two neural network models: a generative model called generator captures the data distribution and output of the generated samples, and a discriminative model that distinguishes whether a sample came from the real dataset or a generated one. During the training procedure, the two models compete with each other. The generator is designed to generate data as realistic as possible so that it is difficult to distinguish them, while the discriminator as a binary classifier aims to identify real and fake data as accurately as possible. The generator and discriminator are optimized alternately until the augmented data are indistinguishable from the actual data.

Figure 5 . Schematic of the generative adversarial network, consisting of two neural networks: a generator and a discriminator. The generator is used to produce the approximated samples from the N -dimensional random noise. The discriminator is used to identify whether a sample is real or fake. These two networks compete with each other and are optimized gradually to realize data augmentation.

Inspired by GAN, a number of new applications have been discovered in terms of images, such as image synthesis, image style transfer, image-to-image translation, and image reconstruction ( Gui J. et al., 2020 ). For optical communication, except for image data, other data types can also be combined with GAN. A network traffic data augmentation technique using GAN was proposed to augment the traffic dataset adaptively for various scenarios ( Li J. et al., 2019 ; Li S. et al., 2019 ). Based on limited experimental traffic data, GAN captured distribution characteristics and then generated massive diverse traffic data, which significantly expanded the training dataset and improved the performance of DL models. Therefore, not limited to image data, GAN can be applied to arbitrary data types by designing appropriate generators for specific application requirements in optical communication.

Deep Reinforcement Learning for Network Automation

Reinforcement (RL) has made great breakthroughs in solving complicated controlling problems based on environment-aware mechanisms. DL plays an important role in perception that can acquire information from observation of the environment and provide current state information, while RL shows powerful advantages in decision-making that can sense complex system states and learn best policies through repeated interactions with the environment, as shown in Figure 6 . DRL combines the perception of DL and the decision of RL to learn a policy that maximizes the cumulative rewards for various tasks, like playing Go, competitive video games, controlling continuous systems in robotics.

Figure 6 . Schematic of deep reinforcement learning combing advantages of perception from deep learning and decision-making from reinforcement learning, to provide a policy for complex controlling problems. Through observation, an agent can acquire the current information from the environment and adjust the action to maximize cumulative rewards for a specific purpose.

The schematic of DRL is displayed in Figure 6 . It can be observed that in DRL there are two main elements (agent and environment) and two core steps (observation and action). The observation provides the current state information of the environment and the action represents the adjustment that the DRL agent makes according to the rewards or punishments from the environment. Therefore, DRL reflects a universal truth that the machine learns from failures in the past and grows after correcting them. Similarly, the agent of DRL learns from rewards and punishments rather than explicit instruction. Through repeated training and learning for a specific purpose, the agent grows powerful gradually to earn more rewards and avoid making mistakes, even exceeding human capacity in many domains.

In the context of optical communication, DRL is particularly useful for network control and automation and thus has been applied in the network layer to automatize the resolutions of routing, resource allocation, orchestration, and configuration ( Chen et al., 2019a , b ; Suárez-Varela et al., 2019 ; Andreoletti et al., 2020 ; Wang et al., 2021 ). A DRL-based routing solution was proposed for the optical transport network (OTN) that can better capture the crucial relationships among the lightpaths and paths in OTN topologies ( Suárez-Varela et al., 2019 ). Considering the real network topologies and traffic profiles, the routing policy learned by the agent outperformed well-known routing heuristics. Moreover, the elastics optical network (EON), where the spectrum distribution becomes extremely flexible and spectrum resource management confronts big challenges ( Yin et al., 2013 ; Zhu et al., 2013 ; Gong and Zhu, 2014 ), requires more automatic and smart control schemes. Accordingly, a DRL-based spectrum assignment scheme was introduced in A DRL-based observer to select the duration of each service cycle adaptively for realizing adaptive and high-quality virtual network function services ( Li B. et al., 2020 ). This study obtained superior results, especially under dynamic, flexible, and complex scenarios.

Additionally, we proposed an adaptive optical transceiver configuration technique using DRL for data center optical networks and passive optical networks ( Li J. et al., 2020 ). The traditional transceivers are only suitable for static scenarios, where the transmission capability is fixed and massive spectrum resources are wasted. Therefore, the flexible optical transceiver is considered as a promising candidate to realize flexible services provisioning but faces the challenges of searching for optimum transceiver parameter sets when considering complex network conditions, including diverse user types, modulation formats, multi-level access distances, quality of transmission, and transmission speed. With the help of DRL, flexible transceivers can be adaptively configured according to network environment states. To improve throughput and spectral efficiency, the agent gradually learns the relationship between network state and the reward of varied configuration actions.

Conclusions

In this paper, powerful DL algorithms were introduced in optical communication to achieve a variety of applications. CNN was used to explore information from image data, including LP mode, OAM mode, eye, constellation, ADTP, AAH, and spectrum diagrams, to implement channel estimation, mode demodulation, optical signal analysis, impairment diagnosis, OPM, DSP, and spectral analysis. RNN was applied to process sequential data, including digital signal waveform, network traffic data, and equipment state parameters, to execute signal pre-distortion and post-compensation, network traffic forecasting, and fault alarming analysis. A data-driven channel modeling scheme was proposed to rethink conventional modeling methods and improve end-to-end learning performance. GAN was adopted to augment image data and sequential data to ensure that the training data were sufficient and diverse. Finally, DRL was introduced to realize self-configuration and the adaptive allocation of optical networks. DL enables optical communication to be more intelligent and adaptive and is expected to make further contributions to optical communication in the near future.

Author Contributions

DW contributed to the study of convolutional neural networks and recurrent neural networks. MZ focused on reinforcement learning-related research.

This work was supported by the National Natural Science Foundation of China (Grant Nos. 61975020 and 61871415), the Key Laboratory Fund (Grant No. 6142104190207), and the Fund of State Key Laboratory of IPOC (BUPT) (Grant No. IPOC2020ZT05), P. R. China.

Conflict of Interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Agrawal, G. P. (2012). Fiber-Optic Communication Systems . Hoboken, NJ: Wiley.

Google Scholar

Andreoletti, D., Velichkova, T., Verticale, G., Tornatore, M., and Giordano, S. (2020). A privacy-preserving reinforcement learning algorithm for multi-domain virtual network embedding. IEEE Trans. Netw. Serv. Manage . 17, 2291–2304. doi: 10.1109/TNSM.2020.3022278

CrossRef Full Text | Google Scholar

Chen, X., Li, B., Proietti, R., Lu, H., Zhu, Z., and Yoo, S. J. B. (2019b). DeepRMSA: a deep reinforcement learning framework for routing, modulation and spectrum assignment in elastic optical networks. J. Lightw. Technol . 37, 4155–4163. doi: 10.1109/JLT.2019.2923615

Chen, X., Proietti, R., and Yoo, S. J. B. (2019a). Building autonomic elastic optical networks with deep reinforcement learning. IEEE Commun. Magaz . 57, 20–26. doi: 10.1109/MCOM.001.1900151

Deligiannidis, S., Mesaritakis, C., and Bogris, A. (2020). Performance and Complexity Evaluation of Recurrent Neural Network Models for Fibre Nonlinear Equalization in Digital Coherent Systems . Brussels: ECOC.

Deng, L. (2014). A tutorial survey of architectures, algorithms, and applications for deep learning. APSIPA Trans. Signal Inf. Process. 3:e5. doi: 10.1017/atsip.2013.9

Fan, Q., Zhou, G., Gui, T., Lu, C., and Lau, A. P. T. (2020). Advancing theoretical understanding and practical performance of signal processing for nonlinear optical communications through machine learning. Nat. Commun . 11:3694. doi: 10.1038/s41467-020-17516-7

PubMed Abstract | CrossRef Full Text | Google Scholar

Gong, L., and Zhu, Z. (2014). Virtual optical network embedding (VONE) over elastic optical networks. J. Lightw. Technol . 32, 450–460. doi: 10.1109/JLT.2013.2294389

Goodfellow, I., Pouget-Abadie, J., Mirza, M., Xu, B., Warde-Farley, D., Ozair, S., et al. (2014). “Generative adversarial nets,” in Proceedings of Advances in Neural Information Processing Systems (New York, NY).

Gu, J., Wang, Z., Kuen, J., Ma, L., Shahroudy, A., Shuai, B., et al. (2018). Recent advances in convolutional neural networks. Pattern Recognit. 77, 354–377. doi: 10.1016/j.patcog.2017.10.013

Gui, J., Sun, Z., Wen, Y., Tao, D., and Ye, J. (2020). A review on generative adversarial networks: algorithms, theory, and applications. arXiv preprint arXiv:2001.06937.

Gui, Y., Wang, D., Guan, L., and Zhang, M. (2020). “Optical network traffic prediction based on graph convolutional neural networks,” in Opto-Electronics and Communications Conference (OECC) (Taipei). doi: 10.1109/OECC48412.2020.9273638

Häger, C., and Pfister, H. D. (2020). Physics-Based Deep learning for fiber-optic communication systems. IEEE J. Select. Areas Commun . 39, 280–294. doi: 10.1109/JSAC.2020.3036950

Karanov, B., Chagnon, M., Aref, V., Lavery, D., Bayvel, P., and Schmalen, L. (2020). “Concept and experimental demonstration of optical IM/DD end-to-end system optimization using a generative model.,” in Optical Fiber Communications Conference and Exhibition (OFC) (San Diego, CA), 48. doi: 10.1364/OFC.2020.Th2A.48

Karanov, B., Chagnon, M., Thouin, F., Eriksson, T. A., Bülow, H., Lavery, D., et al. (2018). End-to-end deep learning of optical fiber communications. IEEE J. Lightw. Technol. 36, 4843–4855. doi: 10.1109/JLT.2018.2865109

Khan, F. N., Fan, Q., Lu, C., and Lau, A. P. T. (2019). An optical communication's perspective on machine learning and its applications. J. Lightw. Technol. 37, 493–516. doi: 10.1109/JLT.2019.2897313

LeCun, Y., Bengio, Y., and Hinton, G. (2015). Deep learning. Nature 521, 436–444. doi: 10.1038/nature14539

Li, B., Lu, W., and Zhu, Z. (2020). Deep-NFVOrch: leveraging deep reinforcement learning to achieve adaptive vNF service chaining in DCI-EONs. J. Opt. Comm. Net . 12, 18–27. doi: 10.1364/JOCN.12.000A18

Li, J., Wang, D., Li, S., Zhang, M., Song, C., and Chen, X. (2019). Deep learning based adaptive sequential data augmentation technique for the optical network traffic synthesis. Opt. Exp. 27, 18831–18847. doi: 10.1364/OE.27.018831

Li, J., Wang, D., Zhang, M., and Cui, S. (2020). Digital twin-enabled self-evolved optical transceiver using deep reinforcement learning. Opt. Lett . 45, 4654–4657. doi: 10.1364/OL.397972

Li, J., Zhang, M., Wang, D., Wu, S., and Zhan, Y. (2018). Joint atmospheric turbulence detection and adaptive demodulation technique using the CNN for the OAM-FSO communication. Opt. Exp . 26, 10494–10508. doi: 10.1364/OE.26.010494

Li, M., Wang, D., Cui, Q., Zhang, Z., Deng, L., and Zhang, M. (2020). “End-to-end learning for optical fiber communication with data-driven channel model, 2020,” in Opto-Electronics and Communications Conference (OECC) (Taipei). doi: 10.1109/OECC48412.2020.9273665

Li, S., Li, J., Zhang, M., Wang, D., Song, C., and Zhen, X. (2019). “Adaptive traffic data augmentation using generative adversarial networks for optical networks,” in Optical Fiber Communications Conference and Exhibition (OFC) (San Diego, CA), 25. doi: 10.1364/OFC.2019.Th2A.25

Lu, P., Zhang, L., Liu, X., Yao, J., and Zhu, Z. (2015). Highly-Efficient data migration and backup for big data applications in elastic optical inter-data-center networks. IEEE Netw . 29, 36–42. doi: 10.1109/MNET.2015.7293303

Lu, X., Lu, C., Yu, W., Qiao, L., Liang, S., Lau, A. P. T., et al. (2019). Memory-controlled deep LSTM neural network post-equalizer used in high-speed PAM VLC system. Opt. Exp . 27:7822–7833. doi: 10.1364/OE.27.007822

Mikolov, T., Karafiát, M., Burget, L., Cernocký, J., and Khudanpur, S (2010). “Recurrent neural network based language model,” in Eleventh Annual Conference of the International Speech Communication Association (Makuhari). doi: 10.1109/ICASSP.2011.5947611

Musumeci, F., Rottondi, C., Nag, A., Macaluso, I., Zibar, D., Ruffini, M., et al. (2018). “An overview on application of machine learning techniques in optical networks. IEEE Commun. Surv. Tutor . 21:13831408. doi: 10.1109/COMST.2018.2880039

Saif, W. S., Esmail, M. A., Ragheb, A. M., Alshawi, T. A., and Alshebeili, S. A. (2020). “Machine learning techniques for optical performance monitoring and modulation format identification: a survey. IEEE Commun. Surv. Tutor. 22, 2804–2821. doi: 10.1109/COMST.2020.3018494

Suárez-Varela, J., Mestres, A., Yu, J., Kuang, L., Feng, H., Cabellos-Aparicio, A., et al. (2019). Routing in optical transport networks with deep reinforcement learning. J. Opt. Comm. Netw . 11, 547–558. doi: 10.1364/JOCN.11.000547

Tschannen, M., Bachem, O., and Lucic, M. (2018). Recent advances in autoencoder-based representation learning. arXiv Preprint arXiv:1812.05069.

Wang, D., Song, Y., Li, J., Qin, J., Yang, T., Zhang, M., et al. (2020). Data-driven optical fiber channel modeling: a deep learning approach. IEEE J. Lightw. Technol. 38, 4730–4743. doi: 10.1109/JLT.2020.2993271

Wang, D., Wang, M., Zhang, M., Zhang, Z., Yang, H., Li, J., et al. (2019). Cost-effective and data size–adaptive OPM at intermediated node using convolutional neural network-based image processor. Opt. Exp. 27, 9403–9419. doi: 10.1364/OE.27.009403

Wang, D., Zhang, M., Li, J., Li, Z., Li, J., Song, C., et al. (2017a). Intelligent constellation diagram analyzer using convolutional neural network-based deep learning. Opt. Exp. 25, 17150–17166. doi: 10.1364/OE.25.017150

Wang, D., Zhang, M., Li, Z., Li, J., Fu, M., Cui, Y., et al. (2017b). Modulation format recognition and OSNR estimation using CNN-based deep learning. IEEE Photon. Technol. Lett . 29, 1667–1670. doi: 10.1109/LPT.2017.2742553

Wang, D., Zhang, Z., Zhang, M., Fu, M., Li, J., Cai, S., et al. (2021). “The role of digital twin in optical communication: fault management, hardware configuration, and transmission simulation,” in IEEE Communications Magazine , 59. doi: 10.1109/MCOM.001.2000727

Wang, Z., Yang, A., Guo, P., and He, P. (2018). OSNR and nonlinear noise power estimation for optical fiber communication systems using LSTM based deep learning technique. Opt. Exp . 26, 21346–21357. doi: 10.1364/OE.26.021346

Yin, Y., Zhang, H., Zhang, M., Xia, M., Zhu, Z., Dahlfort, S., et al. (2013). Spectral and spatial 2D fragmentation-aware routing and spectrum assignment algorithms in elastic optical networks. J. Opt. Commun. Netw . 5, A100–A106. doi: 10.1364/JOCN.5.00A100

Zhang, C., Wang, D., Wang, L., Song, J., Liu, S., Li, J., et al. (2020). Temporal data-driven failure prognostics using BiGRU for optical networks. J. Opt. Commun. Netw . 12, 277–287. doi: 10.1364/JOCN.390727

Zhao, Y., Chen, X., Yang, T., Wang, L., Wang, D., Zhang, Z., et al. (2020). Low-Complexity fiber nonlinearity impairments compensation enabled by simple recurrent neural network with time memory. IEEE Access 8, 160995–161004. doi: 10.1109/ACCESS.2020.3021146

Zheng, H., Lin, F., Feng, X., and Chen, Y. (2020). “A hybrid deep learning model with attention-based conv-LSTM networks for short-term traffic flow prediction,” in IEEE Transactions on Intelligent Transportation Systems . doi: 10.1109/TITS.2020.2997352

Zhu, Z., Lu, W., Zhang, L., and Ansari, N. (2013). Dynamic service provisioning in elastic optical networks with hybrid single-/multi-path routing. IEEE J. Lightw. Technol . 31, 15–22. doi: 10.1109/JLT.2012.2227683

Zia, T., and Zahid, U. (2019). Long short-term memory recurrent neural network architectures for urdu acoustic modeling. Int. J. Speech Technol. 22, 21–30. doi: 10.1007/s10772-018-09573-7

Keywords: artificial intelligence, machine learning, deep learning, optical communications, optical networks

Citation: Wang D and Zhang M (2021) Artificial Intelligence in Optical Communications: From Machine Learning to Deep Learning. Front. Comms. Net. 2:656786. doi: 10.3389/frcmn.2021.656786

Received: 21 January 2021; Accepted: 08 March 2021; Published: 31 March 2021.

Reviewed by:

Copyright © 2021 Wang and Zhang. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY) . The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

*Correspondence: Danshi Wang, danshi_wang@bupt.edu.cn ; Min Zhang, mzhang@bupt.edu.cn

This article is part of the Research Topic

Machine Learning Applications in Optical and Wireless Communication Systems and Networks

Information

• Author Services

Initiatives

You are accessing a machine-readable page. In order to be human-readable, please install an RSS reader.

All articles published by MDPI are made immediately available worldwide under an open access license. No special permission is required to reuse all or part of the article published by MDPI, including figures and tables. For articles published under an open access Creative Common CC BY license, any part of the article may be reused without permission provided that the original article is clearly cited. For more information, please refer to https://www.mdpi.com/openaccess .

Feature papers represent the most advanced research with significant potential for high impact in the field. A Feature Paper should be a substantial original Article that involves several techniques or approaches, provides an outlook for future research directions and describes possible research applications.

Feature papers are submitted upon individual invitation or recommendation by the scientific editors and must receive positive feedback from the reviewers.

Editor’s Choice articles are based on recommendations by the scientific editors of MDPI journals from around the world. Editors select a small number of articles recently published in the journal that they believe will be particularly interesting to readers, or important in the respective research area. The aim is to provide a snapshot of some of the most exciting work published in the various research areas of the journal.

Original Submission Date Received: .

• Active Journals
• Find a Journal
• Proceedings Series
• For Authors
• For Reviewers
• For Editors
• For Librarians
• For Publishers
• For Societies
• For Conference Organizers
• Open Access Policy
• Institutional Open Access Program
• Special Issues Guidelines
• Editorial Process
• Research and Publication Ethics
• Article Processing Charges
• Testimonials
• Preprints.org
• SciProfiles
• Encyclopedia

Article Menu

• Subscribe SciFeed
• Recommended Articles
• Author Biographies
• Google Scholar
• on Google Scholar
• Table of Contents

Find support for a specific problem in the support section of our website.

Please let us know what you think of our products and services.

Visit our dedicated information section to learn more about MDPI.

JSmol Viewer

A study on the irradiance scintillation characteristics of monochromatic led-based visible light communication systems in weak-to-strong turbulence.

1. Introduction

2. propagation of monochromatic led beams with spatially incoherent property in weak-to-strong turbulence, 2.1. irradiance scintillation index in weak turbulence, 2.2. irradiance scintillation index in weak-to-strong turbulence, 2.3. fading probability, 3. numerical results and discussions, 4. conclusions, author contributions, institutional review board statement, informed consent statement, data availability statement, conflicts of interest.

• Tanaka, Y.; Komine, T.; Haruyama, S.; Nakagawa, M. Indoor visible light data transmission system utilizing white LED lights. IEICE Trans. Commun. 2003 , 86 , 2440–2454. [ Google Scholar ]
• Chi, N. LED-Based Visible Light Communications ; Springer: Berlin, Germany, 2018. [ Google Scholar ]
• Shen, C.; Ma, C.; Li, D.; Hu, J.; Li, G.; Zou, P.; Chi, N. High-speed visible laser light communication: Devices, systems and applications. In Broadband Access Communication Technologies XV ; SPIE: San Diego, CA, USA, 2021; Volume 11711, pp. 18–27. [ Google Scholar ]
• Andrews, L.C.; Phillips, R.L. Laser Beam Propagation through Random Media ; SPIE Press: Bellingham, WA, USA, 2005. [ Google Scholar ]
• Komine, T.; Nakagawa, M. Fundamental analysis for visible-light communication system using LED lights. IEEE Trans. Consum. Electron. 2004 , 50 , 100–107. [ Google Scholar ] [ CrossRef ]
• Yamazato, T.; Takai, I.; Okada, H.; Fujii, T.; Yendo, T.; Arai, S.; Kawahito, S. Image-sensor-based visible light communication for automotive applications. IEEE Commun. Mag. 2014 , 52 , 88–97. [ Google Scholar ] [ CrossRef ]
• Shi, M.; Wang, C.; Li, G.; Liu, Y.; Wang, K.; Chi, N. A 5Gb/s 2*2 MIMO Real-Time Visible Light Communication System Based on Silicon Substrate LEDs. In Proceedings of the 2019 Global LIFI Congress (GLC), Paris, France, 12–13 June 2019; pp. 1–5. [ Google Scholar ]
• Wu, T.; Ma, J.; Yuan, R.; Su, P.; Cheng, J. Single-scatter model for short-range ultraviolet communication in a narrow beam case. IEEE Photonics Technol. Lett. 2019 , 31 , 265–268. [ Google Scholar ] [ CrossRef ]
• Eso, E.; Ghassemlooy, Z.; Zvanovec, S.; Sathian, J.; Abadi, M.M.; Younus, O.I. Performance of vehicular visible light communications under the effects of atmospheric turbulence with aperture averaging. Sensors 2021 , 21 , 2751. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Nguyen, Q.D.; Nguyen, N.H. Mobile Application for Visible Light Communication Systems: An Approach for Indoor Positioning. Photonics 2024 , 11 , 293. [ Google Scholar ] [ CrossRef ]
• Shen, T.; Guo, J.; Liang, H.; Li, Y.; Li, K.; Dai, Y.; Ai, Y. Research on a Blue–Green LED Communication System Based on an Underwater Mobile Robot. Photonics 2023 , 10 , 1238. [ Google Scholar ] [ CrossRef ]
• Apolo, J.A.; Ortega, B.; Almenar, V. Hybrid POF/VLC Links Based on a Single LED for Indoor Communications. Photonics 2021 , 8 , 254. [ Google Scholar ] [ CrossRef ]
• Geng, Z.; Khan, F.N.; Guan, X.; Dong, Y. Advances in Visible Light Communication Technologies and Applications. Photonics 2022 , 9 , 893. [ Google Scholar ] [ CrossRef ]
• Yahia, S.; Meraihi, Y.; Ramdane-Cherif, A.; Gabis, A.B.; Acheli, D.; Guan, H. A survey of channel modeling techniques for visible light communications. J. Netw. Comput. Appl. 2021 , 194 , 103206. [ Google Scholar ] [ CrossRef ]
• Álvarez-Roa, C.; Álvarez-Roa, M.; Raddo, T.R.; Jurado-Navas, A.; Castillo-Vázquez, M. Cooperative Terrestrial–Underwater FSO System: Design and Performance Analysis. Photonics 2024 , 11 , 58. [ Google Scholar ] [ CrossRef ]
• Korotkova, O.; Toselli, I. Non-classic atmospheric optical turbulence. Appl. Sci. 2021 , 11 , 8487. [ Google Scholar ] [ CrossRef ]
• Manzoor, H.U.; Manzoor, S.; Manzoor, T.; Khan, T.; Hussain, A. Improved transmission length in the presences of ambient noise and scintillation effect using duobinary modulation in 40 Gbps free space optical channel. Microw. Opt. Technol. Lett. 2020 , 62 , 3163–3169. [ Google Scholar ] [ CrossRef ]
• Ding, H.; Chen, G.; Majumdar, A.K.; Sadler, B.M.; Xu, Z. Turbulence modeling for non-line-of-sight ultraviolet scattering channels. In Atmospheric Propagation VIII ; SPIE: Orlando, FL, USA, 2011; Volume 8038. [ Google Scholar ]
• Jurado-Navas, A.; Raddo, T.R.; Garrido-Balsells, J.M.; Borges, B.H.V.; Olmos, J.J.V.; Monroy, I.T. Hybrid optical CDMA-FSO communications network under spatially correlated gamma-gamma scintillation. Opt. Express 2024 , 15 , 16799–16814. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Liao, L.; Li, Z.; Lang, T.; Chen, G. UV LED array based NLOS UV turbulence channel modeling and experimental verification. Opt. Express 2023 , 17 , 21825–21835. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Ishimaru, A. Wave Propagation and Scattering in Random Media ; Academic Press: New York, NY, USA, 1978. [ Google Scholar ]
• Fante, R.L. Intensity fluctuations of an optical wave in a turbulent medium effect of source coherence. Opt. Acta Int. J. Opt. 1981 , 28 , 1203–1207. [ Google Scholar ] [ CrossRef ]
• Wu, S.; Hu, X.; Han, Y.; Wu, X.; Su, C.; Luo, T.; Li, X. Measurement and analysis of atmospheric optical turbulence in Lhasa based on thermosonde. J. Atmos. Sol.-Terr. Phys. 2020 , 201 , 105241. [ Google Scholar ] [ CrossRef ]
• Tyson, R.K. Adaptive optics and ground-to-space laser communications. Appl. Opt. 1996 , 35 , 3640–3646. [ Google Scholar ] [ CrossRef ]
• Dirkx, D.; Noomen, R.; Prochazka, I.; Bauer, S.; Vermeersen, L.L.A. Influence of atmospheric turbulence on planetary transceiver laser ranging. Adv. Space Res. 2014 , 54 , 2349–2370. [ Google Scholar ] [ CrossRef ]
• Jakeman, E.; Pusey, P.N. Significance of K distributions in scattering experiments. Phys. Rev. Lett. 1978 , 40 , 546. [ Google Scholar ] [ CrossRef ]
• Yi, X.; Liu, Z.; Yue, P. Optical scintillations and fade statistics for FSO communications through moderate-to-strong non-Kolmogorov turbulence. Opt. Laser Technol. 2013 , 47 , 199–207. [ Google Scholar ] [ CrossRef ]
• Al-Habashm, A.; Andrewsl, C.; Phillipsr, L. Mathematical model for the irradiance probability density function of a laser beam propagating through turbulent media. Opt. Eng. 2001 , 40 , 1554–1562. [ Google Scholar ] [ CrossRef ]

Click here to enlarge figure

Share and Cite

Ji, Y.; Chen, W.; Wang, D.; Cheng, C. A Study on the Irradiance Scintillation Characteristics of Monochromatic LED-Based Visible Light Communication Systems in Weak-to-Strong Turbulence. Photonics 2024 , 11 , 567. https://doi.org/10.3390/photonics11060567

Ji Y, Chen W, Wang D, Cheng C. A Study on the Irradiance Scintillation Characteristics of Monochromatic LED-Based Visible Light Communication Systems in Weak-to-Strong Turbulence. Photonics . 2024; 11(6):567. https://doi.org/10.3390/photonics11060567

Ji, Yao, Wensheng Chen, Danning Wang, and Chen Cheng. 2024. "A Study on the Irradiance Scintillation Characteristics of Monochromatic LED-Based Visible Light Communication Systems in Weak-to-Strong Turbulence" Photonics 11, no. 6: 567. https://doi.org/10.3390/photonics11060567

Article Metrics

Article access statistics, further information, mdpi initiatives, follow mdpi.

Subscribe to receive issue release notifications and newsletters from MDPI journals

Design of Optical Fiber Communication Experiments using Simulation Software

Proceedings of the 3rd International Conference on Advances in Science & Technology (ICAST) 2020

3 Pages Posted: 6 Apr 2020

Parshwa Dama

University of Mumbai - K. J. Somaiya Institute of Engineering and Information Technology (KJSIEIT)

Akshay Shinde

Hrishikesh maniyar, zainuddin parkar, anamika singh.

Date Written: April 8, 2020

-Digital fiber optic communication systems address modulation and detection methods for high spectral efficiency and cogency against transmission impairments and are an inherent part of modern communication systems. This paper outlines the design of an open-source tool to perform optical fiber experiments on MATLAB, promoting a greater understanding of optical fiber communication. The tool can be extended by adding experiments. Setting up optical communication systems may be expensive, making its installation in educational institutes unlikely. This tool provides a free alternative to deliver a similar experience to students, enabling possible widespread learning of optical systems in educational institutions. The proposed objective of the project is to design and analyze the simulation model of various modes of optical fiber communication systems using MATLAB and visualize and interact with it using Graphical User Interfaces.

Keywords: optical communication, transmission

Suggested Citation: Suggested Citation

Parshwa Dama (Contact Author)

University of mumbai - k. j. somaiya institute of engineering and information technology (kjsieit) ( email ).

Somaiya Ayurvihar Complex Eastern Express Highway Mumbai, MA Maharashtra 400022 India

Do you have a job opening that you would like to promote on SSRN?

Paper statistics, related ejournals, electrical engineering ejournal.

Subscribe to this fee journal for more curated articles on this topic

Energy Engineering eJournal

Engineering education ejournal, materials science education ejournal.

Academia.edu no longer supports Internet Explorer.

To browse Academia.edu and the wider internet faster and more securely, please take a few seconds to  upgrade your browser .

•  We're Hiring!
•  Help Center

Optical Communication

• Most Cited Papers
• Most Downloaded Papers
• Newest Papers
• Save to Library
• Last »
• Optical Engineering Follow Following
• Optical Measurement Methods Follow Following
• Optical Spectroscopy Follow Following
• Optical Fiber Communications Follow Following
• Optoelectronics & Biomedical Optics Follow Following
• Silicon Photonics Follow Following
• Ultrafast Spectroscopy Follow Following
• Laser optics • Nano Photonics • Optical memory storage Follow Following
• Display Devices Follow Following
• Optics Follow Following

Enter the email address you signed up with and we'll email you a reset link.

• Academia.edu Publishing
•   We're Hiring!
•   Help Center
• Find new research papers in:
• Health Sciences
• Earth Sciences
• Cognitive Science
• Mathematics
• Computer Science
• Academia ©2024

Physical Review Research

• Collections
• Editorial Team
• Open Access

Mitigation of channel tampering attacks in continuous-variable quantum key distribution

S. p. kish, c. thapa, m. sayat, h. suzuki, j. pieprzyk, and s. camtepe, phys. rev. research 6 , 023301 – published 20 june 2024.

• No Citing Articles
• INTRODUCTION
• PARAMETER ESTIMATION IN CONTINUOUS…
• QUANTUM SECURITY ANALYSIS
• CHANNEL TAMPERING ATTACKS
• MACHINE LEARNING CLASSIFICATION RESULTS
• POSTSELECTION OF CLASSIFIED ATTACK
• CONCLUSIONS
• ACKNOWLEDGMENTS

Despite significant advancements in continuous-variable quantum key distribution (CV-QKD), practical CV-QKD systems can be compromised by various attacks. Consequently, identifying new attack vectors and countermeasures for CV-QKD implementations is important for the continued robustness of CV-QKD. In particular, as CV-QKD relies on a public quantum channel, vulnerability to communication disruption persists from potential adversaries employing denial-of-service (DoS) attacks. Inspired by DoS attacks, this paper introduces a threat in CV-QKD called the channel amplification (CA) attack, wherein Eve manipulates the communication channel through amplification. We specifically model this attack in a CV-QKD optical fiber setup. To counter this threat, we propose a detection and mitigation strategy. Detection involves a machine learning (ML) model based on a decision tree classifier, classifying various channel tampering attacks, including CA and DoS attacks. For mitigation, Bob, postselects quadrature data by classifying the attack type and frequency. Our ML model exhibits high accuracy in distinguishing and categorizing these attacks. The CA attack's impact on the secret key rate (SKR) is explored concerning Eve's location and the relative intensity noise of the local oscillator (LO). The proposed mitigation strategy improves the attacked SKR for CA attacks and, in some cases, for hybrid CA-DoS attacks. Our study marks an application of both ML classification and postselection in this context. These findings are important for enhancing the robustness of CV-QKD systems against emerging threats on the channel.

• Received 1 February 2024
• Revised 18 April 2024
• Accepted 8 June 2024

DOI: https://doi.org/10.1103/PhysRevResearch.6.023301

Published by the American Physical Society under the terms of the Creative Commons Attribution 4.0 International license. Further distribution of this work must maintain attribution to the author(s) and the published article's title, journal citation, and DOI.

Published by the American Physical Society

Physics Subject Headings (PhySH)

• Research Areas

Authors & Affiliations

• 1 Data61, CSIRO , Marsfield, NSW 2122, Australia
• 2 Department of Physics, The University of Auckland , Auckland 1010, New Zealand
• * Contact author: [email protected]

Article Text

Vol. 6, Iss. 2 — June - August 2024

Subject Areas

• Quantum Physics

Authorization Required

Other options.

• Buy Article »
• Find an Institution with the Article »

Download & Share

An implementation of Gaussian modulated coherent state QKD protocol with channel tampering attacks in optical fibre. Alice splits a coherent laser source into a quantum signal and LO with an unbalanced beamsplitter. Alice modulates coherent signals with Gaussian distribution N ( 0 , V A ) . The signal and LO are multiplexed and sent through an optical fibre with transmittance T and excess noise ξ . The channel tampering attack modifies the channel parameters up to a distance of D Eve and subsequently, leaves the rest of the optical fibre length of D Bob unchanged up to Bob. Bob measures the power fluctuations P out of the LO in addition to the coherent detector measurement.

Decision tree classifier for classification of channel tampering attacks. The supervised ML model learns from the labeled data the type of channel tampering attack that occurred. It makes a decision based on the features of the data.

The confusion matrix for classifying the types of channel tampering attacks. In (a)  D = 40 km where Eve performs a CA attack at D Eve = 10 km and the σ RIN,LO = 0.01 . The attack parameters for the CA attack are g = 1.12 and p = 1 , for the CA-DoS attack g = 1.12 and p = 0.94 , and for the DoS attack g = 0.9 and p = 0.9 . In (b), the same parameters as (a) but σ RIN,LO = 0.1 . In (c)  D = 40 km where Eve performs a CA attack at D Eve = 1 km and the σ RIN,LO = 0.01 . The attack parameters for the CA attack are g = 1.01 and p = 1 , for the CA-DoS attack g = 1.01 and p = 0.99 , and for the DoS attack g = 0.9 and p = 0.9 . (d) is the same as (c) but with σ RIN,LO = 0.1 .

In (a) is the SKR (in bits per pulse) under attack (left), and with the mitigation (right) in a 40 -km optical fibre. In (b) is the change of the SKR after the CA attack (left) and the SKR improvement (right) plotted against the RIN of the LO and the length of Eve's optical fibre. The block size is N = 10 12 , ɛ cor = ɛ PE = ɛ h = ɛ s = 10 − 9 , p ec = 0.99 , and the number of symbols used for parameter estimation is m = N / 10 . The magenta dashed line is when K Attack = 0 . V A is optimized to maximize K 0 . (Other parameters: f attack = 0.5 , β = 0.9 , v el = 0.05 , η = 0.9 , ξ B = 0.01 , and N 0 = 1 .)

SKR as a function of f attack for K Attack (blue) and K PS (red) with D Eve = 39 km and σ RIN,LO = 0.098 .

CA-DoS attacks with parameters the same as Fig.  4 .

CA attacks with N = 10 8 block size and all other parameters the same as Fig.  4 .

Sign up to receive regular email alerts from Physical Review Research

Reuse & Permissions

It is not necessary to obtain permission to reuse this article or its components as it is available under the terms of the Creative Commons Attribution 4.0 International license. This license permits unrestricted use, distribution, and reproduction in any medium, provided attribution to the author(s) and the published article's title, journal citation, and DOI are maintained. Please note that some figures may have been included with permission from other third parties. It is your responsibility to obtain the proper permission from the rights holder directly for these figures.

• Forgot your username/password?
• Create an account

Article Lookup

Paste a citation or doi, enter a citation.

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

• View all journals
• Explore content
• About the journal
• Publish with us
• Sign up for alerts
• Open access
• Published: 05 June 2024

Skin-inspired, sensory robots for electronic implants

• Lin Zhang 1 ,
• Sicheng Xing   ORCID: orcid.org/0009-0004-8227-2217 2 ,
• Haifeng Yin 3 ,
• Hannah Weisbecker   ORCID: orcid.org/0000-0002-3893-4891 4 ,
• Hiep Thanh Tran 2 ,
• Ziheng Guo 5 ,
• Tianhong Han 6 ,
• Yihang Wang 1 ,
• Yihan Liu 1 ,
• Yizhang Wu   ORCID: orcid.org/0000-0001-6244-8458 1 ,
• Wanrong Xie 1 ,
• Chuqi Huang 1 ,
• Wei Luo 2 ,
• Michael Demaesschalck 5 ,
• Collin McKinney   ORCID: orcid.org/0000-0002-7345-8325 5 ,
• Samuel Hankley 5 ,
• Amber Huang   ORCID: orcid.org/0009-0007-4950-5762 4 ,
• Brynn Brusseau 4 ,
• Jett Messenger 7 ,
• Yici Zou 4 &
• Wubin Bai   ORCID: orcid.org/0000-0003-2872-5559 1  

Nature Communications volume  15 , Article number:  4777 ( 2024 ) Cite this article

4025 Accesses

118 Altmetric

Metrics details

• Biomedical engineering
• Gels and hydrogels
• Mechanical engineering
• Sensors and biosensors

Drawing inspiration from cohesive integration of skeletal muscles and sensory skins in vertebrate animals, we present a design strategy of soft robots, primarily consisting of an electronic skin (e-skin) and an artificial muscle. These robots integrate multifunctional sensing and on-demand actuation into a biocompatible platform using an in-situ solution-based method. They feature biomimetic designs that enable adaptive motions and stress-free contact with tissues, supported by a battery-free wireless module for untethered operation. Demonstrations range from a robotic cuff for detecting blood pressure, to a robotic gripper for tracking bladder volume, an ingestible robot for pH sensing and on-site drug delivery, and a robotic patch for quantifying cardiac function and delivering electrotherapy, highlighting the application versatilities and potentials of the bio-inspired soft robots. Our designs establish a universal strategy with a broad range of sensing and responsive materials, to form integrated soft robots for medical technology and beyond.

Similar content being viewed by others

Skin-inspired soft bioelectronic materials, devices and systems

Stretchable pumps for soft machines

A universal interface for plug-and-play assembly of stretchable devices

Introduction.

The dynamically changing environments drive living organisms to evolve toward inseparable integration of motor and sensor functions 1 , 2 , 3 . Especially, coherent integration between skeletal muscles and sensory skins in vertebrate animals enables their rational and well-organized cooperation orchestrated by neural systems to execute perceptive actions with intelligence. A diverse variety of receptors (mechano, thermal, pain, and others) embedded in the soft skin gathers and encodes tactile data, which not only guides muscular motions to the optimum but also interprets the environment for enhanced awareness and cognition 4 , 5 , 6 , 7 , 8 . Such motor-sensor integration established in the biological systems inspires development of intelligent robotic systems mimicking skin softness to safely explore and interact with dynamic, unstructured, and often uncertain environments, particularly when robots interface with biological tissues and organs to enable precision therapeutics 9 , 10 , 11 , 12 . However, existing robots often lack a seamless integration among actuators, sensors, and controllers, that naturally preserve physical softness and biocompatibility 13 , 14 .

Creating such bio-inspired somatosensory soft robots as implants holds promising potential to innovate medical technology, especially in surgery, diagnosis, drug delivery, prostheses, artificial organs, and tissue-mimicking active simulators for rehabilitation 15 , 16 , 17 , 18 , 19 . Conceptually distinct types of soft robotic implants take the form of shape-morphing and functionalization, are capable of compliance matching biological tissues, retrieving their functional signature, and offering therapeutic treatments 12 , 15 , 16 , 20 , 21 , 22 , 23 . For example, an integrated bladder system consisting of interdigitated capacitive sensors capable of continuous bladder volume detection, and a shape memory alloy-based actuator with strong emptying force for urine voiding, allows for real-time bladder control 20 . A soft gripper based on a shape memory polymer with integration of silver nanowires and a crack-based strain sensor enables conformable contact with a carotid of swine model and measuring its blood pressure 21 . Taking inspiration from the ventricular teeth of hookworm, thera-grippers made of a metal-polymer hybrid actuator and a drug-eluting patch can latch onto the mucosal tissue inside gastrointestinal GI lumen and extend drug release 22 . The combination of sensing and actuation not only enhances diagnostic and/or therapeutic precision for implants via dynamically modulating the structural interface to targeted tissues, but also enables possibility to become artificial organs that offer both needed structural transformation and physiological functions (e.g., electrical signaling, and hormone secretion) 19 , 22 , 24 . Despite the promising progress in soft robotic implants, grand challenges remain in designing materials and manufacturing technologies, to leverage multi-faceted requirements of device performance, including compliant mechanics to match tissue softness, biocompatibility of constituent materials to ensure implantation safety, structural adaptability to prolong device longevity, and biomimicry to enhance device functionality 13 , 14 , 20 , 25 .

Here we present concepts and device designs to achieve untethered, soft robots that follow a biomimicry integration of actuators, sensors, and stimulators, to enable structural adaption and dynamic reconfiguration that minimize tissue damage during implant deployment, release stress at biotic-abiotic interface, increase biocompatibility, and enhance device multi-modal performance with spatiotemporal precision (Fig.  1 and Fig.  S1 ). Demonstrated examples of utilizing such bio-inspired robots include: (i) a robotic gripper that wraps around a bladder to enable coordinated, closed-loop operation of bladder volume evaluation and electrical stimulation to treat underactive bladder, (ii) a robotic cuff that can enclose around a blood vessel for measuring blood flow and pressure, (iii) an ingestible robot that can expand when arriving in a stomach for prolonged monitoring and drug delivery, and (iv) a robotic patch that can actively grasp and release a beating heart for epicardial sensing and electrical stimulation (E-stim), which collectively highlight potential impacts of the bio-inspired robotic designs that naturally integrate actuation, sensing and stimulation within a coherent entity as next-generation electronic implants with physical intelligence.

A A robotic cuff for vascular system. The twisting motion provides physical enclosing around a blood vessel for precise detection of blood pressure and structural support. B A robotic patch for epicardial interface. The gripping motion enables gentle contact with a beating heart without residual straining, to provide real-time quantification of cardiac contractility and temperature, and electrotherapy. C An ingestible robot for digestive system. This structural transition from the shape of a miniaturized pill to a 3D expanded hoop enables extended stay inside stomach to provide both pH sensing and drug delivery. D A robotic gripper for bladder control. The adaptive motion of gripping onto a bladder provides precise tracking of bladder volume and targeted stimulation for treating urological disorders.

The soft robots primarily consist of two integrated, functional layers that emulate relations between sensory skin and underlying muscles. Specifically, one layer represents an electronic skin (e-skin), that is made of functional nanocomposites predominantly based on an in situ solution-based fabrication approach. The other layer represents an artificial muscle, that is based on poly( N -isopropylacrylamide) (PNIPAM) hydrogel, which can reversibly contract and relax upon activation trigger. Significantly, hydrogels, known for their exceptional softness, low activation temperature and nonfibrotic biocompatibility, are generally preferred over other stimuli-responsive materials in implantable applications 26 , 27 , 28 , 29 . The bilayer design composed of the e-skin and artificial muscle represents a heterogeneous configuration with a variety of responsiveness upon exposure to environmental stimuli, which orchestrates its robotic motion. Our in situ solution-based method successfully embeds multiple sensing materials (e.g., silver nanowires ~AgNWs, reduced graphene oxide ~RGO, MXene and poly(3,4-ethylenedioxythiophene) polystyrene sulfonate ~PEDOT:PSS) into a polymer matrix (e.g., polyimide ~PI, and polydimethylsiloxane ~PDMS), enabling the e-skin a versatile platform that highly mimics the skin with complex receptors, and accurately detects the external signals. These functionalities encompass touch, pressure, temperature, and chemical sensing, surpassing the integrative complexity and heterogeneity achievable with 3D printing or other conventional approaches 30 , 31 , 32 . Moreover, inspired by nature (e.g., starfish and chiral seedpods), we can vary designs of the soft robots, enabling various motions (e.g., bending, expanding, and twisting), and corresponding 3D deformed configurations. In addition, the soft robots allow both on-demand transformation and local-region actuation via embedded control circuits, which further increases structural versatility and capability. Moreover, the soft robots can move, sense, and communicate in a wireless closed-loop fashion via integration of control module and data analytics, enabling minimally invasive operations with safe and stable access to enclosed small spaces inside human body.

Bio-inspired multi-modal sensory soft robot

The integrated architecture between the skin and skeletal muscles enables safe and closed-loop interactions with surrounding environment, bringing a feeling of touch and acting of motion seamlessly together in space and time 33 , 34 , 35 . A feature of particular interest in skin is the mechanoreceptors localized at the interface between the epidermis and dermis of skin, which are responsible for detecting a variety of mechanical stimuli, including the fast-adapting (FA) receptors that respond to dynamic forces and slow-adapting (SA) receptors that respond to static pressures (Fig.  2A ) 36 , 37 , 38 . By mimicking the hierarchical architectures of skin and muscle with associated biological functions, our soft robots integrate multi-electronic modules and thermally actuatable hydrogels, realizing both receptor-like sensing functions to detect various stimuli, and on-demand muscle-like contraction to generate physically adaptive motion, from a single integrated platform, endowing soft robots with intelligence in navigating through real-world environments autonomously (Fig.  2B ). Figure  2C shows a fabricated multi-modal sensory soft robot with geometry emulating a starfish. The robot consists of three primary layers: a flexible nanocomposite layer as a multi-modal electronic-skin (e-skin) embedded with distinct sensors (strain, pressure, pH, and temperature) and stimulators (thermal and electrical), a thermally responsive hydrogel layer as an artificial muscle generating actuation force, and a thin bio-adhesive layer as a cushioning medium to form interfacial binding between the e-skin and artificial muscle. Our approach for fabricating the flexible nanocomposite layer can be generally applicable to a wide variety of soft materials and nanomaterials heterogeneously composited within a single matrix material, which enables the potential to form highly integrated systems with a broad range of sensors and stimulators. Specifically, here we demonstrate (1) a thermal sensor made of a nanocomposite of reduced graphene oxide (RGO) and polyimide (PI), (2) a strain sensor made of a nanocomposite of silver nanowires (AgNWs) and polydimethylsiloxane (PDMS), (3) sensing and stimulation electrodes made of a nanocomposite of poly(3,4-ethylenedioxythiophene): polystyrene sulfonate (PEDOT:PSS) and PI, and (4) a thermal heater made of a nanocomposite of AgNWs and PI. The flexibility and versatility of this strategy allow the as-fabricated functional units to be easily integrated into each arm of the starfish robot in a monolithic fashion, as illustrated in Fig.  2D , realizing a sensory skin for the robot to enable environmental awareness. Figure  2E displays an example of an ultrathin multi-modal e-skin equipped with six nanocomposite sensors (The detailed fabrication process appears in Fig.  S2A .). The e-skin is further encapsulated with a parylene layer (thickness ~2 μm) to enhance its durability during prolonged applications 39 , 40 . This is subsequently attached onto a piece of predesigned thermally responsive PNIPAM hydrogel that serves as an artificial muscle for the soft robot (Fig.  S2B ). The PNIPAM hydrogel can undergo a dramatic volumetric reduction of about 90% as the temperature shifts from 25 °C to 60 °C. Notably, this significant and rapid deswelling behavior is initiated only when temperature is beyond its lower critical solution temperature (LCST 32–34 °C), enabling versatile actuation capabilities within biological environments (Fig.  S3A, B , and Fig.  2H ) 41 , 42 . Additionally, we can adjust the LCST by incorporating acrylamide (AAm) into the PNIPAM hydrogel to form poly(NIPAM-co-acrylamide) (P(NIPAM-AAM)), enabling us to tailor the LCST to align with varying application scenarios that may require operations at distinct temperature ranges (Fig.  S4 ) 43 , 44 , 45 . The as-fabricated soft robot can actively form a conformal and stress-free interface with curvilinear biological surfaces, indicating its inherent mechanical softness and high biocompatibility. This adaptability minimizes potential risks related to mechanical incompatibility, facilitating its smooth integration with targeted tissues/organs (Fig.  2F, G , and Fig.  S3C ). At an elevated temperature (40 °C), the robot consisting of the stimuli-responsive artificial muscle (PNIPAM layer) and non-stimuli-responsive e-skin (multi-modal layer) tends to bend toward the muscle side on the basis of asymmetrical responsive properties (Fig.  2H , Fig.  S5A–C ) 17 , 46 .

A Schematic illustration of epidermis-dermis-muscle structure of skin. B Bio-inspired structure of soft robot from skin. C Conceptual illustration of the integrated multi-modal sensory soft robot with distinct nanocomposite sensors functionalized into each arm. D Schematic illustration of a starfish-inspired multi-modal sensory soft robot. Left: An exploded view highlighting 3 primary constituent layers, including a flexible multi-modal layer, a bio-adhesive layer, and an actuation hydrogel layer. Right: Schematic illustration highlighting the multi-material integration within the multi-modal layer including: (i) Silver nanowires (AgNWs) and polyimide (PI) as a flexible heater; (ii) AgNWs and PDMS as a strain sensor; (iii) Reduced graphene oxide (RGO) and PI as a temperature sensor; (iv) Poly(3,4-ethylenedioxythiophene) polystyrene sulfonate (PEDOT:PSS) and PI as sensing and stimulation electrodes. E Optical image of the flexible e-skin with six nanocomposite sensors. Optical image showing conformal attachment of the soft sensory robot onto human skin ( F ) and porcine tissue ( G ) with high mechanical compliance. H Top: Schematic showing a temperature-responsive bending from a bilayer of poly(N-isopropylacrylamide) (PNIPAM) and polyimide-based nanocomposite. Bottom: Volumetric shrinkage of PNIPAM across a temperature range of 25–60 °C. The data points represent the mean value from n  = 3 independent experiments, and the error bars are in S.D. I – K Optical images and corresponding finite element modeling of thermal-triggered structural reconfiguration of soft robots. Colors indicate von Mises stress magnitude. I Biomimicry soft gripper encloses upon heating at 40 °C. J Chiral seedpod-inspired robot reverses helix at 40 °C. K Soft robotic pill expands upon heating at 40 °C. L , M Schematic and optical images of anisotropic integration of various functional materials into a polymeric matrix to form a multi-modal sensing system using an in situ solution-based approach. L A RGO/PI-based temperature sensor and an AgNW/PI-based heater on the same polyimide side. M PEDOT:PSS/PI-based electrodes and RGO/PI temperature sensors integrated on the two opposite sides of a PI layer, respectively. Scale bars, 5 mm.

Figure  2I , Fig.  S5D , and Supplementary Movie  S1 provide an example of a robotic starfish reversibly closing and opening its rays under a temperature shift between 23 °C and 40 °C. Furthermore, such multi-layer integration allows a diverse collection of robotic designs that undergo various types of actuations. Fig.  S6 , 7 and Supplementary Movie  S2 highlight soft sensory robots featuring starfish-like structures with an arbitrary number of rays (e.g., four and six) and a fishbone-like structure with four pins. In addition, tuning the design layout of the multi-layer structure can yield complex deformation and structural reconfiguration beyond simple bending motion. Here, inspired by the self-twisting of chiral seedpods, we constructed a geometry with parallel strips in the e-skin layer (Additional details appear in Fig.  S8A ) 47 , 48 . Upon thermal stimulation, the PNIPAM hydrogel contracts while the PI maintains stability, generating a differential contraction across the structure. The differential thermal response induces a tilting torque, leading to local saddle-like curvature and twisting motion of the integrated robotic systems (Fig.  2J , Fig.  S8B and Supplementary Movie  S3 ). Our findings closely align with the simulation results as shown in Fig.  S8C . We also develop a soft robotic pill based on an anchored hydrogel/nanocomposite tri-layer structure, where a hydrogel-muscle layer is surrounded by two separate e-skin layers that bond to the muscle layer at the edges (Fig.  S9A shows the detailed structure design.). Upon thermal actuation, the pill can self-expand into a 3D ring shape, since the contraction of the muscle layer drives the e-skin layers to buckle out of the plane, as shown in Fig.  2K , Fig.  S9B , and Supplementary Movie  S4 . Furthermore, compared with a make-and-transfer method 30 , 31 , 32 , our in situ solution-based fabrication approach enables the integration of sensors into the e-skin matrix in a single step, enhancing mechanical and electrical performance by reducing interfacial resistances and improving mechanical conformity, thereby significantly improving sensitivity and responsiveness. As aforementioned, this approach also offers a versatile platform that can be constructed using a broad range of functional nanomaterials hybridized with a polymeric matrix to form a multi-modal sensing system. By selecting materials that offer unique functional attributes, from AgNWs known for their conductivity and flexibility to graphene and MXene for their high surface area, electrical conductivity, and hydrophilicity 49 , 50 , 51 , 52 , this system provides a possible emulation to the skin that contains complex somatosensory units, where various mechanoreceptors and thermoreceptors distributed in the epidermal and dermal layers enable the spatiotemporal recognition of the magnitude and location of touch and temperature stimuli 53 , 54 . Fig.  S10 provides a representative example of in situ integration of AgNW/PDMS-based strain sensor, in which the strain sensor has a good response to the stretching. Figure  2L shows the proposed solution-based approach enables a freestanding PI film to integrate multi-functional modalities including an RGO/PI-based temperature sensor and an AgNW/PI-based heater (Fig.  S11A ), endowing soft robots with both thermal sensing and stimulation. Moreover, Fig.  2M displays a more complicated integration paradigm with multi-layer stacking, where different electronic components (e.g., PEDOT:PSS/PI-based conductive electrodes and RGO/PI temperature sensors) can be distributed in different layers of the e-skin to achieve simultaneous functional versatility and compactness. This assembly technique ensures the e-skin remarkable thinness and flexibility, enhancing its effective performance for implantable applications (Fig.  S11B ). The X-ray photoelectron spectroscopy (XPS) characterization on the e-skin layers reveals the precise nanoscale integration of active materials within a polymer matrix, as detailed in Figs.  S12 – S14 and Supplementary Note  S1 . It showcases the optimal distribution and intermolecular bonding of the composite components, effectively addressing the common challenge of uneven dispersion of nanomaterials, which usually undermines the performance of conventional composites. Our approach minimizes the polymer amount required to integrate nanomaterials into composite functional modules and utilizes excess polymer as an insulating layer to separate modules, preventing interference between their electrical and chemical signals, thereby ensuring that each functional module operates independently and effectively. This simple approach combines the distinct properties of each constituent, achieving a balance between structural integrity and functional versatility 55 , 56 , 57 , 58 , 59 . This advanced level of integration would be of great value for soft robots that seek to achieve multifunctionality and local sensing capabilities approaching skin.

On-demand robotic actuation with spatiotemporal control

Programmable stimuli-responsive soft robotic systems capable of working in enclosed or confined spaces and adapting functions under changing situations hold great promise as next-generation medical robots 60 , 61 . The realization of versatile morphing modes through local-actuation control is crucial for enhancing on-demand actuation. Here, we develop soft robots with cognitive capabilities via unifying programmable actuation and in situ sensing. Figure  3A shows a sensory robotic arm primarily consisting of a PNIPAM-hydrogel-based muscle layer and a multi-modal e-skin layer that is designed with an AgNW/PI nanocomposite actuation heater and a RGO/PI nanocomposite thermal sensor. Fig.  S15A shows layer-by-layer stacking as a simple and effective approach for fabricating the e-skin. This approach stands out for its capability to fabricate multi-layered e-skin integrating diverse functionalities within an unified e-skin framework, offering a sophisticated level of device customization. Figure  3B and Fig.  S15B demonstrate that the multifunctional nanocomposite film is highly flexible and can be tightly bonded onto the hydrogel layer via n-butyl cyanoacrylate adhesive. Importantly, the bio-adhesive layer exhibits robust adhesion under both dry and wet conditions, as displayed in Fig.  S16 . This feature ensures the reliability of implantable devices in the dynamic and moist human body environment. When the liquid PI is cast onto the nanomaterial film (e.g., AgNWs and RGO), the liquid PI penetrates into the interconnected pores of the three-dimensional (3D) network, owing to the low viscosity and low surface energy of the liquid PI 62 . The scanning electron microscope (SEM) images, including both top and side views as shown in Fig.  3C, D , Fig.  S17A–F , and Fig.  S21A–C , present that the curing process fully buries all the nanomaterials inside the PI matrix, resulting in a uniform composite free from observable voids. This prevents separation of the nanomaterials from the polymeric matrix, thus minimizing untended side effects to the neighboring tissues. The Fourier-transform infrared spectroscopy (FTIR) and X-ray powder diffraction (XRD) results (Fig.  S17G, H , Fig.  S21D, E , and Supplementary Note  S2 ) also confirm the chemical structure of AgNW/PI and RGO/PI nanocomposites suggesting a successful fabrication of high-quality nanocomposites of functional materials and polymer matrix 63 , 64 , 65 , 66 . Figure  3E demonstrates that the resultant AgNW/PI nanocomposite conductor is highly conductive, twistable, and bendable, and thus used as a highly flexible heater for electrothermal actuation. Thermal images in Fig.  3E show the infrared thermograph of a Joule-heated e-skin in the resting state and under both bending and twisting conditions. The electrical heater distributed inside the e-skin exhibits a stable and uniform temperature distribution without degradation of the temperature level at the deformed points, ensuring effective thermal transport from the heater to the hydrogel-based muscle layer. Figure  3F and Fig.  S18A depict the transient electrothermal response of the flexible heater applied by various powers in an ex vivo environment. The saturation temperature of AgNW/PI nanocomposite heater increases with the supplied power as more Joule heat is generated, and the lower critical solution temperature (LCST) of PNIPAM hydrogel at 34 °C can be obtained at low input power (<~1 W) (Fig.  S18A ). As shown in Fig.  S18B , the flexible electrical heater exhibits steady heating and cooling cycles, indicating high repeatability and remarkable heating stability using AgNW/PI nanocomposite. Figure  3G shows varying the amount of electric power applied to the heater effectively modulates the resultant bending angle θ (as defined in Fig.  S19A ) of the robotic arm, thus realizing precise, on-demand access to intermediate morphologies along the bending pathway. Figure  3G also indicates that the actuator reaches its peak deformation upon achieving thermal equilibrium, and importantly, this maximum bend is maintained as long as there are no changes in thermal conditions. We further evaluated the mechanical force generated by the soft robotic finger which incorporates a PNIPAM hydrogel layer roughly 1 mm thick, under various input powers. Fig.  S19B shows that the static force exhibits a noticeable increase with rising temperature. At a temperature of 40 °C, the force reaches a maximum of 32 mN. Additionally, it is observed that the generated force remains consistent throughout 40 cycles of alternating power on and off (0.35 W), indicating the robust reversibility of the soft robot (Fig.  S19C ). When compared to similar hydrogel-based soft actuators, our design consistently achieves a relatively high output force, as shown in Table  S1 67 , 68 , 69 , 70 , 71 , 72 , 73 , 74 . Moreover, our study demonstrates that the thickness of the hydrogel layer determines the actuation performance of the device. As depicted in Fig.  S20A , thicker hydrogel layers induce more pronounced shape morphing at a set activation temperature, while a slower response due to their increased mass. The thicker layers are also capable of generating higher actuation force, thereby enhancing the actuation capability of the device (Fig.  S20B ). However, the increased device volume may limit its applicability in minimally invasive implantable devices, where compactness is a key factor. In the device configuration, embedded sensors are strategically positioned within a single, thin e-skin layer, thus minimizing the required thickness of the hydrogel layer for efficient actuation and allowing the integrated system to be relative compact (Fig.  S20C ). Our fabrication technique accommodates customization of the hydrogel layer thickness to optimize device performance for specific applications, demonstrating the method’s flexibility and adaptability to various biomedical needs. The utilization of a RGO/PI-nanocomposite temperature sensor enables real-time tracking of environmental temperature during muscle motion. The in situ solution processing ensures RGO network is uniformly distributed in PI matrix to provide precise pattern registration with good electrical conductivity (~400 S/m) and stable performance in aspects of electrical linearity, sensitivity, and repeatability. Figure  3H illustrates the resistive change in a relatively linear relation with temperature for the RGO/PI thermal sensor. The temperature coefficient of the resistance (TCR) of the RGO/PI thermal sensor is >0.5%/°C, featuring its high thermal sensitivity. On the other hand, the RGO/PI-based thermal sensor exhibits a stable performance after 1000 bending cycles, and even after immersing in PBS solution. Figure  3I and Fig.  S21F show performance of the thermal sensor in response to cycles of temperature rise and drop, indicating good sensing stability. In addition, Fig.  S21G shows consistent measurements of RGO/PI nanocomposite sensing performance in comparison with a commercial thermal resistor (ERT-J0ET102H), indicating excellent sensing accuracy. To exemplify this versatility, we have also successfully fabricated MXene-based sensors, as illustrated in Fig.  3J, K and Fig.  S22 . MXene-based materials possess inherent biocompatibility, rendering them ideally suited for incorporation into implantable biomedical devices without eliciting adverse reactions or compromising the surrounding biological environment 52 , 75 .

Illustration of an exploded view ( A ) and optical image ( B ) of a soft robotic arm comprising a PNIPAM hydrogel layer, a RGO/PI thermal sensor, and an AgNW/PI actuation heater, bonded using n-butyl cyanoacrylate adhesive. Joule heating from the AgNW/PI heater triggers bending of the robotic arm. Scale bar, 5 mm. C , D SEM images of nanocomposite film surfaces. Here, deep reactive ion etching (DRIE) of partial regions reveals the anisotropic integration within the films, distinguishing pristine PI from AgNW/PI ( C ) and RGO/PI regions ( D ). AgNWs and RGO, with higher surface-free energy than PI, are uniformly dispersed inside the PI matrix, enhancing wetting and binding properties. Scale bars, ( C ) 1.5 μm, ( D ) 10 μm. E Infrared thermograph of an AgNW/PI heater demonstrating consistent heating performance even after 1000 bending and twisting cycles. F Surface temperature of this heater varies with input electric power, functioning efficiently at low power. G The bending angle of a soft robotic arm changes with varying input electric power. H Resistive response of the RGO/PI thermal sensor during bending, twisting, and PBS immersion, across temperatures from 23 °C to 92 °C. Data points represent mean values from n  = 3 independent experiments, error bars in S.D. I Static cycling test of the RGO/PI thermal sensor. Here, the left y-axis represents resistance changes, while the right y-axis shows corresponding temperature changes. J Resistive response of an MXene-based thermal sensor across temperatures from 23 °C to 55 °C. K Temperature measurement comparison between the MXene/PI thermal sensor and a commercial thermal resistor (ERT-J0ET102H). L Stepwise coiling actuation of a soft robotic probe via sequential thermal stimulus. Top shows infrared thermograph and bottom shows structural changes from a flat to coiled state. Scale bars, 10 mm. M Integrated RGO/PI thermal sensors enable temperature measurement of localized regions for proprioceptive sensing. N Optical images demonstrating on-demand motion control of a three-arm soft robotic gripper via sequentially programming input power. Scale bars, 5 mm. 3 measurements are repeated with similar results to ( C ), ( D ), ( I ), ( J ), ( K ), ( L ).

Well-controlled shape morphability with high spatiotemporal resolution is essential for soft robots to execute complicated tasks safely in biological environments 76 , 77 . Here, the integration of e-skin and artificial muscle allows the bio-inspired soft robots to realize motions of local regions independently, which collectively can lead to a wide range of shape morphability. Figure  3L demonstrates the locally controlled morphability with a soft robotic finger, of which the e-skin includes a series of AgNW/PI-based heaters for localized thermal activation, and RGO/PI-based thermal sensors for temperature monitoring (Fig.  S23A ). As shown in Fig.  3L and Fig.  S23B , the robotic finger undergoes a stepwise self-coiling motion via a sequential stimulus and perceives its local temperature simultaneously by the embedded thermal sensors (Fig.  3M ). More complex actuation modes or configurations are accessible owing to the combined ease of processing and ability to introduce multifunctional units. Figure  3N presents a three-arm soft robotic gripper where each arm contains various types of sensors, including an optical sensor, a thermal sensor, and a strain sensor, respectively, and an electrical heater (Fig.  S24A ), to enable independent control of motion for each arm, thus allowing on-demand gripping and interaction with targeted objects (Supplementary Movie  S5 ). Here, the observed variability in bending configuration can be attributed to differences in the positioning and depth of the functional modules within the e-skin layer. Moreover, the soft robotic systems can be structurally tailored to support a range of motions. For example, Fig.  S24B, C highlights a soft starfish-like robot with four sensory arms, and a soft robotic cuff with optoelectronic sensors, respectively. The electrothermal stimulus along with distributed sensing capabilities enables programmed actuation not only on demand but also regulated simultaneously by the sensing feedback (e.g., Supplementary Movie  S6 ).

Furthermore, our soft robotic system exemplifies advanced sensory-motor integration, leveraging the synergistic relationship between embedded sensors and actuators to achieve dynamic adaptivity and responsiveness to environmental changes. A prime example is a temperature-sensitive control system, as shown in Fig.  S25A , which utilizes real-time sensory feedback to dynamically adjust heating in response to environmental temperature changes. The operational principle, as detailed in Fig.  S25B and Supplementary Note  S3 , involves a microcontroller-driven algorithm that interprets temperature input collected by a resistive temperature sensor, and modulates the electric heater’s current accordingly, enabling rapid adaptations to achieve and maintain a preset temperature. Fig.  S26 presents a soft robotic finger’s real-time response to temperature variations, ensuring stable shape adaptation through this regulatory mechanism. Moreover, this intelligent control significantly improves safety by preventing the risk of overheating, thereby ensuring the system’s safe operation in various thermal conditions, highlighting our device’s ability to provide precise thermal management, enhancing both efficacy and safety in its applications 78 , 79 .

Wireless sensing and actuation of soft sensory robot

Wireless operation of actuation as well as therapeutic and/or diagnostic functions is essential for implantable robots to minimize tissue damage and implant infection 76 , 80 , 81 , 82 . However, existing schemes for wireless operation of robots mostly require sophisticated circuitry for energy harvesting and storage, which may introduce non-negligible heat (~80 °C), unfavored space occupation, system complexity, limited lifetimes, and high cost 83 , 84 . To overcome the preceding challenges of actuation/sensing integration in soft implantable robots, we report a bio-inspired soft robot as a representative example to demonstrate remote, battery-free operation and communication in both sensing and actuation.

Figure  4A–C show schematic illustrations of the system design consisting of sensor and actuator components. Our strategy for wireless sensing relies on a passive inductor-capacitor (LC) resonance circuit formed by a planar inductor coil and a parallel plate based on capacitor polyacrylamide (PAAm) hydrogel, as illustrated in Fig.  4B . A change in a biomechanical event (e.g., vessel filling, cardiac contraction/relaxion, and bladder filling/voiding) can lead to a corresponding change in capacitance of the hydrogel-based capacitor, which can be quantified by recording shift in resonance frequency ( f s ) of the LC circuit according to the equation \({f}_{s}=1/2\pi \sqrt{{LC}}\) , where L and C are the inductance and capacitance of the resonance circuit, respectively 85 , 86 . The inductor coil couples to an alternating electromagnetic field through a readout probe, enabling quantitative measurement of the input return loss (S11) using a vector network analyzer (VNA) (Fig.  4D and Supplementary Note  S4 ). The proposed PAAm-based sensor with relatively low modulus, intrinsic stretchability, and biocompatibility, can detect the pressure through variation of capacitance between the two electrodes. Figure  4F shows a linear correlation between the measured ΔC/C 0 and applied pressure. The slope of the linear fitting curve reports gauge factor (GF) of the sensor reaching ~3%/kPa in the clinically relevant range of pressure (0–4 kPa). Figure  4G&H shows the resonant frequency decreases from 340 MHz to 260 MHz in response to the applied pressure decreasing from 0 to 3.5 kPa, and the pressure sensitivity of the sensor is ~26.7 kHz/kPa.

A Schematic of a soft sensory robot featuring an electrical heater, a polyacrylamide (PAAm)-based pressure sensor, and two inductive coils for transmission sensing signals ( B ) and electrical power ( C ). B Exploded view of the sensing components containing a capacitor with two electrodes, a PAAm-hydrogel dielectric layer, and a copper (Cu) inductive communication coil. C Exploded view of actuation components consisting of a PNIPAM actuation hydrogel, a flexible electrical heater, and a radiofrequency (RF) power harvester based on a copper coil. D Equivalent circuit diagram of wireless pressure sensing, where pressure variations alter the capacitance and, thus, resonance frequency which is captured wirelessly through a vector network analyzer (VNA). E Equivalent circuit diagram of wireless actuation, where a transmitting coil connected to an RF power amplifier energizes the receiving coil, powering the heater for robotic motion. F Measured capacitive change of the PAAm-based pressure sensor in response to applied pressure. G Measured shift of resonance curves of the PAAm-based pressure sensor in response to applied pressure. H Change of the LC resonant frequency as a function of applied pressure serving as a signal-transduction scheme for wireless pressure detection. I Thermal distribution during wirelessly harvesting energy exhibits minimal heating in the receiving coil and efficient power consumed by the electrical heater, ensuring minimal heat damage to surrounding bio-environments. J Optical images of a soft sensory robot undergoing a wireless actuation to transform from a flat state to a bent state. K The output power as a function of frequency, optimized at ~15 MHz. L The output electrical power varies with the input power. M The temperature change of the electrical heater over time under various output powers used for wireless actuation. N Optical images of the deformed RF coils including bending, twisting, and distorting. O Measured resonance frequency of RF coil under various shape deformations. Here the bending angle θ is 60 o . Scale bars, 5 mm. Data are presented ( F ), ( K ), ( L ) from n  = 3 independent experiments, and the error bars are in S.D.

Moreover, wireless electromagnetic power transmission is an attractive solution to overcome limitations imposed by implantable systems with discrete batteries, especially for devices operating in enclosed places such as the human body 86 , 87 , 88 . Here, we demonstrate the bio-inspired, sensory robot allows for electrically controlled locomotion in this wireless, battery-free manner (Fig.  4A ). Figure  4C shows layout of the wireless actuation design including a PNIPAM actuation hydrogel, a flexible electrical heater, and a radio-frequency (RF) power harvester based on a Cu coil. The RF harvester is made of a triple-layer structure (Fig.  S27A–C ). Magnetic coupling between the transmitting and receiving coil generates electric currents to the heater that thermally stimulates the artificial muscle (Fig.  4E & Fig.  S27D ). As shown in Fig.  4J and Fig.  S28 , the robotic finger can undergo a bending motion upon receipt of electromagnetic energy from a power-transfer module. Figure  4I shows an infrared image of the robotic finger highlighting concentrated heat on the electrical heater and minimum heat on the receiving inductor, which collectively ensures a stable power supply and minimum heat damage to surrounding bio-environments in potential usage as implants. Figure  4K shows the output power as a function of frequency. The wireless power transfer system achieves the highest harvested power ~1.05 W at a frequency of 15 MHz. Figure  4L indicates the harvested power can be tuned via varying the input power at the optimal frequency 15 MHz, and the power efficiency reaches around 80%, which is sufficient to trigger the heater for raising local temperature to drive motions of the soft robot (Fig.  4M ).

The soft robotic system features spatiotemporal locomotion wirelessly controlled by RF signal modulation. To demonstrate the working principle of this method, we fabricate a soft robotic gripper with three arms as an illustrative example. This strategy employs a three-lead LC receiver coil designed based on the frequency-response characteristics of magnetic resonance coupling (Fig.  S29A, B ), allowing for the formation of two different configurations with distinct inductive values. These configurations are paired with corresponding capacitors to form LC circuits with different resonant frequencies (Fig.  S29C ). By coupling with different transmission coils operating at distinct transmission frequencies, the coils can selectively deliver power to specific heaters, thereby enabling the activation of individual robotic arms (Fig.  S29D–F ). Furthermore, we conduct a further investigation into the influence of shape deformation, including bending, twisting, and distorting, on the transfer performance of the RF harvester (Fig.  4N ). Fig.  S27E demonstrates that the zeros of the reactance of the receiving circuit remain constant throughout the entire process, indicating minimal disruption to the coil’s resonance frequency 89 , 90 . Additionally, Fig.  4K and Fig.  4O exhibit a reduction in transmission efficiency resulting from decreased mutual inductance between the transmission and the receiving coils due to shape deformation. For all the deformations shown here, the resonance frequency of the coil remains within an acceptable range for power harvesting, and the power transmitted remains above the minimum level required for the robotic actuation. Supplementary Note  S5 and Figs.  S30 – S32 further demonstrate the effect of various design and operational parameters on the power transfer system 91 , 92 , 93 . This yields systematic metrics for designing a well-tuned WPT system, capable of fulfilling the specifications of the targeted application, concurrently decreasing power expenditure, and evading possible risks to living organisms 78 , 94 . Such wireless robotic systems may serve as a promising solution to safe, real-time monitoring of internal pressure needed for various medical procedures.

Soft sensory robots interfacing with various internal organs

Soft sensory robots have significant potential for medical-device applications that warrant safe, synergistic interaction with humans 95 , 96 . Here, we develop robots with the capability to actively morph into 3D configurations to generate a stress-free and stable interface with targeted organs for enhanced sensing, stimulation, and drug delivery. In vitro tests with artificial organ models demonstrate the versatilities and potential applications of our soft robots.

Urinary bladder dysfunction is one of the emerging issues in an aging society, which not only leads to loss of voluntary control over the bladder muscles, but also cuts off sensorial feedback to central nervous system 97 . In most cases of bladder dysfunction, the patients are not able to sense the fullness of bladder with urination, making it a challenge to time the action for voiding treatment (e.g., electrical stimulation) 98 . Therefore, realizing the voiding treatment in a timely manner requires a monitoring system that measures the bladder status continuously.

Here, we develop a soft robotic gripper for both real-time assessment of bladder volume and voiding treatment in a wireless closed-loop control fashion. The as-prepared robotic system, illustrated in Fig.  5A–C and Fig.  S33A, B , consists of a flexible hydrogel-based actuator, a 3D buckled strain sensor, an electrical stimulator, and a control module. The actuator includes a passive layer made of a patterned Au/PI bilayer as an electrical heater, and an active layer of PNIPAM hydrogel. Upon an electrical trigger delivered via an inductive coil, the robotic gripper can bend and wrap around the bladder conformally and gently (Fig.  5A ) to ensure precise measurement of bladder volume and minimize stress at the interface. The strain sensor integrated into the robot can detect bladder volume continuously. Here, the strain sensor includes an elastic PAAm hydrogel film and a serpentine Au/PI resistor to form a buckled 3D structure for enhanced sensitivity. The fabrication steps for the buckled sensor appear in Materials and Methods. We use a balloon model to mimic the natural bladder behavior of filling and emptying to validate performance of the integrated soft robotic gripper. Fig.  S33C shows the sensor conformally attaches onto the balloon surface with the biocompatible adhesiveness of PAAm hydrogel 99 . The softness of PAAm hydrogel in the robot achieves minimal strain disruption to the bladder during the device operation. Injecting and extracting water into the balloon with a syringe pump realizes the behavior of bladder filling and emptying (Fig.  5D ). The change of sensor resistance exhibits a strong linear correlation with bladder volume during its expansion and shrinkage (Fig.  5E and Fig.  S33D, E ), thus serving as an indicator in bladder-volume control. Figure  5F demonstrates the system repeatability in real-time monitoring during multiple cycles of bladder filling and emptying.

A A fully implantable, soft robotic gripper precisely measuring bladder volume and providing electrical stimulation through wireless closed-loop control. B The control platform including a wireless power harvesting network, a full bridge amplifier, a voltage regulator, and signal conditioning circuits integrated with a Bluetooth System-on-Chip, and a MOSFET switch. C Block diagram for the bladder stimulation module. D A soft robotic gripper deployed onto an artificial bladder, demonstrating deflation and inflation cycles. E Measured resistive characteristics of the buckled strain sensor on the artificial bladder correlated with volume changes. F The 3D buckling strain sensor monitoring real-time volumetric changes during cyclic bladder operations, with resistance and volume changes displayed on dual axes. G Programmed electrical stimulation (top) and measured volume of an artificial bladder (middle and bottom). This demonstration is conducted using a volume threshold of ~100 mL, and amplitude of 3 V. A slight delay in the deactivation process could be attributed to the microcontroller unit’s response time. H A soft robotic cuff enclosing around a blood vessel for monitoring blood pressure. I Optical image of the cuff on an artificial vessel stimulating blood circulation. J Measured resistive changes of the strain sensor at various simulated blood pressures. K Fluidic pressure measurement in an artificial artery system using the soft robotic cuff, displaying resistance changes on the left y-axis and pressure changes on the right y-axis. L A soft ingestible robot designed for continuous stomach pH monitoring and extended drug delivery. M Optical images showing the robot entering, expanding and blocking in the stomach. N Electrical response of PEDOT:PSS/PVA hydrogel to pH change ranging from 3 to 7 over time, with resistance on the left y-axis and pH changes on the right y-axis. O Rhodamine-B embedded into the poly lactic-co-glycolic acid (PLGA) matrix to form a drug delivery patch concealed inside the robot. Its release is measured by UV-vis absorbance over an hour at different temperatures. Scale bars, 5 mm. Data are presented ( E ), ( J ) from n  = 3 independent experiments, and the error bars are in S.D.

The readout of the bladder volume is achieved through a voltage divider circuit (VDC) consisting of a reference resistor connected in series with the as-fabricated strain sensor. This configuration has minimal impact on the wireless power transmission between coils (Fig.  S34 ). To achieve a closed-loop control of electrical stimulation in the treatment of a dysfunctional bladder system, we further program the Bluetooth-Low-Energy (BLE) System-on-Chip (SoC) (BLE SoC) to enable a pulse-width modulation (PWM) instance. The instance, together with on-board power amplification and filtering circuits, facilitates the application of programmable electrical stimulation in response to an increased strain in the bladder. This enables on-demand electrotherapy and closed-loop control of the robotic implant, as illustrated in Fig.  5B ,  C . A demonstration of the control scheme can be found in Fig.  S35 and Fig.  5G . When the balloon’s volume reaches a predetermined threshold, set here at 100 mL, the control system initiates electrical stimulation. After successful voiding below the threshold, the system automatically deactivates the stimulation. While electrical stimulation has shown promising results in enhancing bladder control in various studies and clinical trials, its efficacy can differ across individuals 100 , 101 . The effectiveness of electrical stimulation for bladder voiding and its required voltage levels requires further investigation beyond the scope of our current study. However, our prototype showcases the potential of integrating sensing and actuation mechanisms to facilitate timely and adaptive interventions for bladder dysfunction.

The on-demand motion provided by the soft robot facilitates device implantation and ensures benign and stable contact with targeted tissue or organs 12 , 102 , 103 . Here we develop a soft robotic cuff that can enclose around a blood vessel upon thermal stimulation (Fig.  5H ), to enable real-time measurement of blood pressure. As shown in Fig.  S36A , the soft robotic cuff consists of a muscle layer based on a PNIPAM and an e-skin layer embedded with a strain sensor based on a serpentine Au ribbon. Notably, the e-skin layer uses a pattern of parallel strips in the PI film, that forms a 45 o angle with the longitudinal direction of the device, which, by coupling with the muscle layers, facilitates formation of a helix structure. An in vitro model of artery that uses a rubbery tube with a pulsatile flow of water to create a simulated pulsation pattern validates performance of the soft robotic cuff (Fig.  S36B–D and Fig.  5I ). The helix formation of the robotic cuff provides a gentle and stable coupling with the artificial artery, of which the measured signal (resistive change) exhibits a linear relation with the inner fluid pressure (Fig.  5J ). Figure  5K demonstrates the capability in continuous measurement that captures patterns of simulated pulsation, further confirming its utility in measuring blood pressure.

Furthermore, such bio-inspired designs of soft sensory robots can also configure into an ingestible platform. Here, we show that a soft ingestible robot is capable of prolonged residence in the stomach for pH monitoring and drug delivery. Fig.  S37A depicts the proposed structure of the ingestible robot based on a slab-shaped muscle layer sandwiched by two e-skin layers consisting of a drug delivery module and a pH sensing module (Fig.  5L , more details appear in the Materials and Methods.). The miniaturized size of the robot facilitates swallowing and transport through the esophagus to stomach (Fig.  5L ). Once in the stomach, the robot self-expands that prevents passage through the pylorus for an extended duration in the gastric environment (Fig.  5M and Fig.  S37B ). The soft pH sensors based on PEDOT:PSS/poly(vinyl alcohol) (PVA) embedded onto the e-skin of the robot provide continuous measurements of gastric pH (Fig.  S37C and Fig.  5N ). The pH sensor exhibits a linear relationship between the resistive change and pH of the fluidic environment within the range of acidic (pH ~ 3) to basic (pH ~ 8) (Fig.  S37E ). To realize prolonged drug release, the integrated drug patches use poly(lactic-co-glycolic acid) (PLGA) as a matrix to load drugs and the robotic motion as a trigger mechanism that exposes the drug-loaded patches to stomach fluid for initiating drug release. As a demonstration, we use a biocompatible dye, rhodamine-B (RB), as a model drug (5 mg of RB per 0.5 g PLGA). Figure  5O and Fig.  S37H show the UV-vis absorption spectra and the corresponding dosages, respectively, of the RB-loaded drug patch, immersed in PBS solution (at pH ~ 5) for 1 h under various temperatures ranging from 25 °C to 45 °C, (Fig.  S37F, G show the calibration curve based on the measured UV-vis absorption spectra.). Such ingestible soft robots highlight a synergistic combination between sensing function and robotic motion to realize on-demand, prolonged control of drug delivery.

Additionally, our device offers versatile adaptability for diverse application scenarios through customizable dimensions, sensor positioning, and geometric layouts, ensuring it aligns with the unique morphologies and functional demands of targeted tissues/organs. This flexibility is essential for enabling minimally invasive deployment. As illustrated in Figs. S38, S39 , Table  S2 , and Supplementary Note  S6 , the design’s adaptability enhances soft robotic technologies for effective integration in a broad spectrum of biomedical applications 104 , 105 , 106 , 107 , 108 , 109 , 110 , 111 . We also explored the bioadhesive behavior of our device on targeted tissues/organs. We observed that hydrogel’s inherent adhesiveness is significantly related to its water content and temperature 112 . As shown in Fig.  S40 , there is a decline in adhesive strength as temperatures approach the hydrogel’s LCST. While this inherent adhesive capability contributes to the initial secure placement of the device, it’s noteworthy that solely relying on this property might not guarantee a durable bond, especially as the hydrogel experiences dehydration. However, this temperature-responsive adhesiveness can play a complementary role in enhancing the device’s grasp by counterbalancing any potential decrease in force due to hydrogel reswelling.

A soft robotic thera-gripper for epicardial sensing and electrical stimulating

Cardiac implants that can monitor and regulate heart rhythms are critical for patients with severe cardiac diseases 113 , 114 . Despite the broad implementation of cardiac implants or pacemakers in clinical settings, however, existing devices are usually rigid, undeformable, and lack structural reconfigurability to adapt to dynamic motions of beating heart, which precludes optimum performance chronically and safely 115 . Here, we report a soft robotic thera-gripper that can gently envelop a heart to perform spatiotemporal monitoring of electrophysiological activity, temperature, and strain, and provide therapeutic capabilities (e.g., electrical stimulating). The soft robotic thera-gripper contains four multi-functional arms, emulating a starfish, to ensure effective latching onto the epicardial surface of which the soft mechanics ensures negligible disruption to cardiac dynamics. Fig.  S41A presents an exploded view of the robotic thera-gripper, which contains actuators based on PNIPAM actuation hydrogel, two temperature sensors made of thermal resistors, two electrical stimulating electrodes made of Au, and four strain sensors made of serpentine Au/PI resistors (The fabrication approach appears in Materials and Methods, and Fig.  S41B .). Figure  6A shows the thera-gripper at its resting state features minimally invasive insertion with a medical catheter, and forms a bowl shape at the actuation state, via a slightly raised temperature, to gently hold a beating heart for a safe, stable interface (Fig.  6C, D ). The design employed PNIPAM hydrogel with a LCST 34 °C that is closely aligned with natural body temperature to achieve necessary shape deformation. The initial heating serves primarily to accelerate the actuation, but after achieving the desired state, continuous electrical heating becomes unnecessary. This feature allows the device to effectively adapt and function within the physiological temperature range without the need for ongoing thermal input. Figure  6B shows the corresponding FEA result of an actuated soft robotic thera-gripper. Moreover, we examined the biocompatibility of the soft robots in vitro and in vivo. Here, mouse 3T3-J-2 cells exposed to complete soft robotic devices with constituent materials including pure PNIPAM hydrogel and functional nanocomposites (e.g., AgNWs, RGO, PEDOT:PSS & MXene), remain robust healthy and maintain stable viability (Fig.  6E, F , Figs.  S42 , S43 ) 116 , 117 . Additionally, histological analysis reveals that the soft robots implanted inside chest chamber of mice induce no observable inflammation or other adverse effects to surrounding tissues, indicating good biocompatibility for long-term operation (Figs.  S44 , 45 ). The incorporated temperature sensors provide feedback information on device temperature that govern the structural transformation upon contact with cardiac tissues (Fig.  6G ). The E-stim electrodes integrated into our thera-gripper can generate electrical impulses to regulate cardiac functionality (Fig.  6H and Fig.  S41C, D ). Fig.  S46 shows measured ECG traces, highlighting effective transmission of electrical impulses (voltage ranges from 500 mV to 2 V with 1 ms width at 2.65 Hz) onto cardiac tissues. The e-skin layer consists of microelectrodes for capturing electrical activity of the heart, which serves as essential guidance in operating electrical stimulation (Fig.  S47 ). Figs.  S48 , S49 showcase the simultaneous sensing and stimulation capabilities on a beating heart with an in vivo mouse model, demonstrating its capability in a broad spectrum of potential therapeutic applications 118 , 119 .

A Schematic illustration showing the thera-gripper features minimally invasive insertion at the resting state and wraps onto the surface of a beating heart at the actuation state. The thera-gripper contains four strain sensors made of serpentine Au/PI resistors, two E-stim electrodes based on Au, and two temperature sensors made of thermal resistors. B Finite element modeling of the actuation state. The colors in the legend indicate the magnitude of the von Mises stress. C Image of a soft robotic thera-gripper grasping on the epicardial surface of a living mouse heart. Scale bar, 5 mm. D Schematic illustration showing the thera-gripper position on a mouse heart, where the strain sensors (labeled as S1, S2, S3, and S4) are located onto different heart chambers for locally monitoring of dysfunctional tissue. E Confocal microscope images of 3T3-J-2 cells before (control) and after exposure to as-prepared soft robots integrated with an e-skin layer and a PNIPAM hydrogel-muscle layer, incubated at 37 °C and 39 °C for 48 h. Scale bars, 50 μm. F Comparative cell viability before and after soft robot’s exposure, indicating that 3T3-J-2 cells exposed with the soft robot and cultured at the elevated temperature (39 °C) have no decreased viability. Scale bars, 50 μm. Mean ± S.D. n  = 3. P value by Unpaired t -test. G Temperature measurements from the thera-gripper during its deployment onto the mouse heart, demonstrate the device’s capability to monitor thermal variations in real-time. H The surface ECG trace during electrical stimulation using a pair of Au E-stim electrodes. Optical images of a healthy heart ( I ) and an injured heart 2 weeks after myocardial infarction (MI) ( J ). The MI area is shown by the white dashed circle in ( J ). M-mode echocardiographic images from a healthy ( K ) and post-MI heart ( L ). Representative measurements of local cardiac contractions before ( M ) and after MI ( N ) using a soft robotic thera-gripper wrapping onto a living mouse heart.

Cardiovascular disease is one of the leading causes of death worldwide affecting more than 17.9 million people per year 120 . Real-time and continuous monitoring of myocardial functions (e.g., contractility) is desired for patients with high risks of cardiac arrest, which not only ensures proactive treatment prior to occurrence of adverse events, but also provides comprehensive evaluations of therapeutic effects during postoperative care 121 , 122 , 123 . A soft thera-gripper, integrated with strain sensors distributed on its arms, can provide continuous, spatially resolved quantification of myocardial strains, holding great potential in precision treatment for cardiac diseases. Here, a chronic model of myocardial infarction (MI) with permanent ligation of the left coronary artery (LCA) of a mouse enables assessment of the sensing capability of a thera-gripper post to its robotic motions to establish optimized sensing interfaces (details appear in Materials and Methods.). Figure  6I, J shows the infarcted area that turns into pale myocardium. Echocardiographic strain imaging provides continuous quantification of regional myocardial function. Fig.  S50A show the short-axis views in echocardiographic imaging (B-mode) from a normal and post-MI left ventricle (LV) of heart, respectively. The M-mode images, as shown in Fig.  6K, L , display the fraction shortening (FS) decreases from 75% to 15% upon occurrence of MI, indicating the heart exhibiting moderate dysfunction accompanied by regional contractility loss (Supplementary Note  S7 and Fig.  S50B ) 124 , 125 . Although echocardiography is an invaluable tool allowing for reliable diagnostic information necessary for the clinical decision, it often lacks convenient accessibility and usually takes more than 30 min to capture the movement and function of the heart muscle and heart valves 126 . Figure  6D and Fig.  S51A show that the multi-strain sensors (labeled as S1, S2, S3, and S4) of the implanted thera-gripper distributed across four chambers of the heart, respectively, highlighting the key advantage of our thera-gripper with real-time strain sensing over traditional imaging tools in recording local contractions continuously and simultaneously. Fig.  S51B shows the response of a representative strain sensor, revealing mechanical rhythms of the cardiac cycles, consistent with ECG recordings. Ligation of coronary artery leads to myocardial infarction that causes ST-segment elevation, as shown in Fig.  S51C . Figure  6M, N shows the contractility patterns of the right and left atria (RA and LA), and the right and left ventricles (RV and LV) under normal and MI conditions, respectively, measured by an implanted thera-gripper where the output features of the sensors are determined by their experienced strain that significantly correlates to their positions on epicardial surface. The S4 (LV) displays the largest amplitude due to a maximum experienced strain, indicating the intrinsic highest myocardium strength in the LV chamber. Figure  6J reveals the area of infarction 2 weeks after the MI surgery. The results in Fig.  6N demonstrate that the LV outflow obstruction changes patterns of ventricular contraction. The sensors S2 (RV) and S4 (LV) experience a reduced strain change due to the consequent loss of contractile myocardium, which, correspondingly, reduces force of myocardial contractility and decreases heart rate. Collectively, such bio-inspired design establishes a foundation for implantable robots to harness on-demand motion for structural adaptation inside body and sensory functions for real-time optimization of therapeutic outcomes.

Our post-implantation evaluation revealed that the hydrogel-based thera-gripper remained intact and securely attached to cardiac tissue as intended, demonstrating its durability and effectiveness over time (Fig.  S52A ). Notably, the E-stim electrodes and thermal sensor maintained optimal performance, effectively delivering electrical stimulation (Fig.  S52B, C ) and precisely sensing temperature fluctuations (Fig.  S52D ) even after a 2-week period. These results support its feasibility for long-term therapeutic and diagnostic applications.

In this study, we report concepts and device designs to achieve untethered soft robots that highly emulate biological systems and seamlessly integrate actuators, sensors, and stimulators to enable structural adaption and reconfigurable interfaces that minimize tissue damage in vivo, enhance mechanical match at biotic-abiotic interface, increase biocompatibility, and improve multimodal functionality with spatiotemporal precision. Our soft robots primarily consist of an e-skin layer made of multi-modal nanocomposites that mimic receptors in biological skin to perceive various external stimuli, and an artificial muscle layer made of thermally responsive PNIPAM hydrogel to generate adaptive motion. We employ an in situ solution-based approach to fabricate the flexible multi-modal e-skin. This facile method represents a versatile platform, for a broad range of functional materials (e.g., AgNWs, RGO, and PEDOT:PSS) to be incorporated into a polymeric matrix (e.g., PDMS and PI) to form various types of sensors (e.g., temperature, pressure, and strain) with high spatiotemporal resolution. Such biomimicry design of soft robots offers versatility in mechanical motions including bending, twisting, and expanding as well as diversity in structural deformation, including configurations resembling starfish, fishbone, chiral seedpod, and others. On-demand actuation triggered by electrothermal stimulation from electrical heaters embedded in the e-skin allows precise, independent control of regional parts of the soft body. Furthermore, integration with wireless modules enables the robots to be controlled and communicated without tethering, even when implanted inside body. To demonstrate the broad utility, we develop soft robots that are tailored for specific application scenarios. Specifically, we fabricate a soft robotic gripper that can wrap around a bladder to enable coordinated, closed-loop operation of bladder-volume evaluation and electrical stimulation to treat an overactive bladder, a robotic cuff that can twist around a blood vessel for measuring blood pressure, and an ingestible robot that can reside in a stomach for prolonged pH sensing and drug delivery. In vivo studies with a mouse model demonstrate capabilities of a soft robotic thera-gripper in gently enveloping a beating heart, spatiotemporal assessment of electrophysiological activity, quantification of cardiac contractility, and supplying electrical stimulation for functional regulation. These demonstrations showcase the potential applications of such soft robots as next-generation biomedical implants with structural intelligence and multi-functionalities. Future advancements could further enhance the synergistic interaction between soft implantable robots and biological tissues, to achieve long-term biocompatibility and stability in dynamic physiological environments for improving treatment of chronic diseases.

N-isopropylacrylamide (NIPAM, 98%) was purchased from TCI. Poly (vinyl alcohol) (PVA, 99 + % hydrolyzed), poly(D,L-lactide-co-glycolide) (Mw 50,000-75,0000), acrylamide (AAm, 99%), N, N’-Methylenebisacrylamide (BIS, 99%), N,N,N’,N’-Tetramethyl-ethylenediamine (TMEDA, 99%), and ammonium persulfate (APS, >98%) were purchased from Sigma-Aldrich. Rhodamine B and silver nitrate (AgNO 3 99.9%) was purchased from Themo Scientific. Ethylene glycol (EG, >99%) was purchased from BDH Chemicals. Polyvinylpyrrolidone (M W  ≈ 55,000, PVP) and poly(3,4-ethylenedioxythiophene)-poly(styrenesulfonate) (PEDOT:PSS, 3.0–4.0%) were purchased from Sigma-Aldrich. Copper chloride (CuCl 2 ) was purchased from Ward’s Science. The graphite powder was purchased from Spectrum Chemical Manufacturing Corp. The sodium nitrate (NaNO 3 , >99%), hydrogen peroxide (H 2 O 2 , 30% w/w), the potassium permanganate (KMnO 4 , >99%), hydrochloric acid (HCl, 36.5–38%) and sulfuric acid (H 2 SO 4 , 95–98%) were purchased from BDH chemicals. Ethylenediamine (EDA, 99%) was purchased from Alfa Aesar.

Materials characterization

The SEM images were taken by the field emission scanning electron microscope (Hitachi S-4700 with EDS). The XPS spectra were obtained by Kratos Axis Supra x-ray photoelectron spectrometer, allowing to determine the elemental composition of the top ~10 nm of the sample surface. The XRD patterns of functional nanocomposites were obtained using the Rigaku SmartLab theta-theta diffractometer. The FTIR spectra were recorded using Hyperion 1000 with Tensor 27 spectrometer. The thermal images were taken with an infrared (IR) camera (FLIR ETS320). The performance of the drug-release patch was characterized with a Ultraviolet-visible (UV-vis) spectrometer (VWR UV-1600PC).

Multi-modal sensory soft robots with bio-inspired designs

The fabrication of thermo-responsive hydrogel: Poly(N-isopropylacrylamide) (PNIPAM) was synthesized based on precipitation polymerization. In a typical example, 500 μL NIPAM as monomer (25 wt%), 50 μL BIS as crosslinker (0.5 wt%), 140 μL APS (1 wt%)/200 μL TEMED (2 wt%) as initiator were mixed to form the precursor solution. After 30 min, the precursors can polymerize, in situ, to form the thermos-responsive PNIPAM hydrogel.

The synthesis of poly(NIPAM-co-acrylamide) (P(NIPAM-AAm)) hydrogel: The synthesis of the PNIPAM-co-PAAm was conducted through a free radical polymerization method. In a typical procedure, a mixture of 450 μL NIPAM monomer (25 wt%) and 50 μL AAm monomer (at concentrations of 5 wt%, 10 wt%, 15 wt%, 20 wt%, and 25 wt%) was prepared. Additionally, 50 μL of BIS crosslinker (0.5 wt%), 140 μL of APS initiator (1 wt%), and 200 μL of TEMED (2 wt%) were added to form the precursor solution. After 30 min, the precursors can polymerize to form the thermos-responsive P(NIPAM-AAm) hydrogel.

The fabrication of AgNWs: AgNWs were synthesized based on a modified polyol method. 100 mL of EG containing NaCl (~0.05 mM), PVP (~189 mM), AgNO 3 (~0.0014 mM), and CuCl 2 (~0.017 mM) were added to a round-bottom flask and heated at 185 °C for 1 h in an oil bath. Then, 30 mL AgNO 3 EG solution (~0.12 M) was added dropwise with vigorous stirring. After the reaction was completed, the flask was cooled to room temperature. The AgNW suspension in the EG was diluted with 30 mL water and sedimented for 14 h. The supernatant was decanted, the water was added to reach 160 mL in total volume of the solution. The suspension was sedimented for 14 h and decanted. Then, 50 mL water was added to this mixture, and 100 mL acetone was slowly added with gentle mixing. After centrifugation, the as-formed AgNW pellets was fully resuspended in a 20 mL PVP aqueous solution (~0.5 wt%). The cleaning process was repeated 4 times. Finally, the AgNW pellets were stored in water for later use.

The fabrication of reduced graphene oxide: The graphene oxide (GO) was synthesized based on a modified Hummers method. In general, 5 g of graphite powder and 2.5 g of NaNO 3 were added into 120 mL sulfuric acid in an ice bath under stirring for 2 h. Subsequently, 15 g of KMnO 4 was slowly added into the solution under a temperature <20 °C. After 1 h, the reaction temperature was raised to 35 °C for overnight reaction. 150 mL H 2 O was added to the mixture for 12 h reaction under 98 °C. Then, 500 mL of H 2 O and 20 mL of H 2 O 2 were added to the mixed solution. Finally, the solution was washed with HCl (1 M) and H 2 O until the pH was natural. The EDA was used as the reducing agent to reduce the graphene oxide.

The fabrication of a soft robot inspired by a starfish: The fabrication process of the multi-modal functional electronic-skin (e-skin) is shown in Fig.  S2A . A polyimide (PI) substrate (thickness ~ 10 µm) was patterned on a pre-cleaned and plasma-pretreated glass slide using laser-cutter (six rectangular, each is 15 mm × 6 mm). The AgNWs solution (50 wt%), the RGO suspension (5 wt%), and PEDOT:PSS solution were drop cast onto the patterned glass slide and heated at 50 °C for drying. After the solution was dried, the structured functional materials were fabricated with laser cutting (SFX-50GS). The thermal effects of laser patterning ensure minimal residual materials remain, as localized heating effectively eliminates any leftovers, facilitating seamless progression to subsequent processing steps. Then, a thin layer of liquid PI was spin-coated onto the patterned thin film of functional nanomaterials and cured at 150 °C for 1 h. Finally, the resultant film was cutted out with a laser beam and peeled off from the glass slide. The as-formed multi-modal e-skin was encapsulated with a layer of parylene (thickness ~ 2 μm) and bonded onto PNIPAM hydrogels via the adhesive glue (3 M Vetbond 1469c) (Fig.  2C ) to generate the multi-modal sensory soft robot with a bio-inspired starfish design. Here, laser hatching parameters for patterning nanomaterials are set: for AgNW/PEDOT:PSS/RGO, an infrared laser power of 12% (50 W) with a hatching speed of 5000 mm/s and frequency of 40 kHz; for Au/MXene, the settings are adjusted to an infrared laser power of 10% (50 W), a hatching speed of 1000 mm/s, and frequency of 40 kHz, ensuring precision in the material’s functional structuring.

The fabrication of a soft robot inspired by chiral seedpods : The gold nanomembrane (Au, thickness ~ 200 nm) and adhesive chromium layer (Cr, thickness ~ 10 nm) were deposited by the magnetron sputtering on the PI film (thickness ~ 10 μm). The patterns for Au electronics and PI substrate shown in Fig.  S8A were formed using the laser cutting machine. Then the e-skin made of patterned Au/PI film and encapsulated by a parylene layer (thickness ~ 2 μm), is attached to the PNIPAM hydrogel (15 mm × 6 mm) using the bioadhesive layer.

The fabrication of a soft robotic pill: Firstly, the PEDOT:PSS solution was dropcast onto a pre-treated glass slide spacer, dried under 50 °C and patterned using a laser-cutter. Secondly, the PEDOT:PSS/PI nanocomposite thin film was formed by spin-coating liquid PI onto the patterned PEDOT:PSS, and fully curing at 150 °C for 1 h. Next, the e-skin layer received a protective coating of parylene, ~2 μm in thickness. Finally, two pieces of PEDOT:PSS/PI nanocomposite film were adhered to a piece of PNIPAM hydrogel at the edges (Fig.  S9A ).

Static finite element analysis for various soft robots

3D finite element analyses (FEA) in commercial software ABAQUS were utilized to predict the shape transformation process of soft robots with different patterns and dimensions. The elastic modulus ( E ) and poison’s ratio ( υ ) used in the simulations were E PNIPAM  = 90 KPa, υ PNIPAM  = 0.30 for PNIPAM hydrogel, and E PI  = 2.5 GPa, υ PI  = 0.34 for PI.

Mechanical characterization

Adhesion force was tested by the standard 180° peel test with the Instron machine (Mark-10 ESM303). All tests were conducted with a constant peeling speed of 13 mm/min.

Mechanical tests were conducted on rectangular-shape specimens with the dimensions of 10 mm in width, 2 mm in thickness, and 16 mm in length, using the Instron machine (Mark-10 ESM303).

The sensory-motor integration within the soft robotic system

The robotic gripper and the external circuitry were connected in series with an NI DMM amperemeter set for DC current measurement. The device was cooled to 22 °C in ambient temperature before the system was started up to capture its response to a sudden decrease in ambient temperature. The temperature readouts recorded by the device’s integrated sensor were logged via a microcontroller unit (MCU) and cross-referenced with data from FLIR thermal camera. Both the current and temperature data were analyzed using custom Python script designed specifically for this purpose.

The fabrication of polyacrylamide (PAAm) hydrogel pressure sensor: In polyacrylamide (PAAm) hydrogel synthesis, 500 μL AAm as monomer (25 wt%), 50 μL BIS as crosslinker (0.5 wt%), 140 μL APS (1 wt%)/200 μL TEMED (2 wt%) as initiator were mixed to form the precursor solution. After 10 min, the precursor can polymerize, in situ, to form the PAAm hydrogel. Then a parallel-plate capacitor-based pressure sensor was formed through sandwiching PAAm hydrogel between two electrodes made of Au/PI bilayer (Fig.  4B ). Here, to mitigate potential stability issues regarding hydrogel swelling and its impact on pressure sensing, we propose two mitigation solutions. Firstly, we incorporate a protective encapsulation layer around the hydrogel. This layer is engineered to be permeable enough to facilitate pressure transmission while simultaneously shielding the hydrogel from direct exposure to body fluids that may induce excessive swelling. Secondly, we refine the hydrogel’s composition by increasing the crosslinker concentration to diminish swelling sensitivity without impairing the hydrogel’s ability to sense pressure. By applying these strategies will enhance the stability of the hydrogel as a pressure sensor for implantable applications.

The fabrication of the soft sensory robot: Laser cutting of the Au/PI bilayer formed the electrical heater that was connected to a Cu receiver coil (Fig.  4C ). This assembly was then covered with a thin layer of parylene (thickness ~ 2 μm). Following this, the PNIPAM muscle was bonded to the electrical heater to form the bilayer structure. Finally, the actuation component was integrated with the PAAm-based pressure sensor (Fig.  4A ). Here, the circuits for sensing and actuating can be integrated together onto a single piece of Au/PI bilayer film (Fig.  S27C ).

The characterization of wireless power transmitting and pressure sensing: The detailed power transfer performance and pressure sensing data acquisition are explained in Supplementary Note  S4 .

Wireless control of locomotion: We implement a three-lead LC receiving network based on the frequency response characteristics of magnetic resonance coupling (Fig.  S29A ). In this system, two leads are connected to the full length of the coil, while a third electrode is connected to the middle of the coil loops, resulting in the formation of a smaller inductor with the common electrode (Fig.  S29B ). The two coil loops with different inductive values are paired with different capacitors (Full length ~ 200 pF; Half length ~ 47 pF) to form LC circuits with different resonant frequencies yet similar small quality factors (Fig.  S29C ). Each soft robotic arm is connected to an individual pair of leads to harvest RF power transmitted at different frequencies.

LCST tunability of PNIPAM-based hydrogel

Measurement of bending angle of soft robots based on PNIPAM-co-PAAm hydrogel: The bending angles were measured as a function of time under different electrical powers in both a simulated in vivo environment (37 °C, PBS solution) and an in vitro condition (room temperature, PBS solution). The measurements were performed for three different levels of AAm incorporation: 0 wt%, 5 wt%, and 10 wt%.

A soft robotic gripper for bladder control

The fabrication of a robotic gripper: The process began with the fabrication of e-skin layer via laser cutting Au/PI bilayer film and applying a parylene coating. The e-skin layer includes electrical heaters, electrical stimulators, and serpentine Au resistors (Fig.  S33A ). Then, the PNIPAM hydrogel was adhered onto the electrical heater, and the Au/PI resistor was anchored onto one piece of PAAm hydrogel at the edges to form a strain sensor with a buckled structure (Fig.  S33C ).

Measurement of biomimetic bladder volume: A balloon model was used to mimic the natural bladder to evaluate the actuation and sensing performance of the soft gripper. After thermal stimulation, the soft gripper can wrap around the balloon and the strain sensor can attach onto the balloon surface via the PAAm hydrogel. A syringe pump was used to inject and extract water into the balloon to imitate the filling and emptying behavior of natural bladder. The PowerLab (Model 16/35, AD Instruments) allowed the recording of output signals.

Wireless sensing of resistive strain sensor and close-looped control of bladder electrical therapy robot: A power harvesting and signal conditioning circuit was fabricated and soldered using the method of soft PCB fabrications (Supplementary Note  S4 ). We developed an alternative approach for powering and signal conditioning. This strategy enables the accurate readout of the resistive strain sensor without significantly affecting the wireless power transmission between coils. Specifically, we fabricate a full-bridge rectifier with surface-mount diodes, enabling the conversion of the alternative current (AC) obtained via the receiver coil into direct current (DC). Subsequently, the rectified DC is fed into a 3.3 V low-dropout (LDO) regulator. The output from this regulator serves as the power source for both a Bluetooth-Low-Energy (BLE) System-on-Chip (SoC) and a voltage divider circuit. The voltage divider circuit consists of a reference resistor connected in series with the as-fabricated resistive strain sensor. The voltage divider circuit consists of a reference resistor connected in series with the as-fabricated resistive strain sensor. The voltage across the resistive strain sensor through the voltage divider circuit was sampled by a 14-bit on-chip Successive Approximation Analog to Digital Converter (SAADC). A custom BLE service transmits the sampling value to host terminals acting as BLE clients and parsers to allow wireless readout of strain values. For achieving a closed-loop control of electrical stimulation in the treatment of dysfunctional bladder system, we further programmed the BLE SoC by enabling a pulse-width modulation (PWM) instance. The instance, combined with on-board power amplification and filtering system, enables the programmed electrical stimulation corresponding to detected bladder strain variations.

A robotic cuff for vascular system

The fabrication of a robotic cuff: Firstly, we used laser cutter to fabricate the e-skin layer with a pattern of parallel strips that exhibit a 45° angle with the longitudinal direction of the device. The e-skin layer contains a serpentine Au ribbon as a strain sensor (Fig.  S36A ). Subsequently, this layer was coated with a parylene film ~2 μm in thickness. Finally, the PNIPAM hydrogel-muscle layer was bonded to the e-skin layer.

Measurement of biomimetic blood pressure: As shown in Fig.  S36B , to mimic the arterial environment, we employed a silicone tube (Transparent Silicone Tube 4 mm ID × 5 mm OD, wall thickness ~ 0.5 mm) with large stretchability and flexibility as an arterial artery. The pulsatile water flow (30–260 mL/min) is generated with two flow rate controllable pumps. Pump 1 maintains a constant flow rate to establish a baseline pressure, while Pump 2, connected to a solenoid valve, is regulated by a relay. This valve opens and closes periodically, replicating the pulsatile pressure of blood flow. The relay is further controlled with a pre-programmed microcontroller. All the parts for the setup were purchased from local venders.

An ingestible robot for digestive system

The fabrication of a drug-releasing patch: In brief, 5 mg poly(lactic-co-glycolic acid) (~PLGA) was dissolved in 5.1 g acetone containing 5 mg Rhodamine B (~RhB). Then the mixture was poured into a poly(dimethylsiloxane) (~PDMS) spacer, and acetone was removed from the mixture under vacuum condition. Finally, the as-formed film of PLGA/RhB was cut into small pieces of drug-release patch with a diameter ~2 mm.

The fabrication of PEDOT:PSS/PVA hydrogel sensor: For hybrid hydrogel fabrication, first a 10 wt% PVA solution was made by dissolving PVA powder in water. Then 5 wt% PEDOT:PSS aqueous dispersion was added into the PVA solution followed by slow mixing for 24 h. The prepared PEDOT:PSS/PVA solution was poured into glass spacer, followed by freezing at −20 °C for 8 h and thawing at 25 °C for 3 h for three times.

The fabrication of a soft ingestible robot: The process began with the formation of e-skin layer. The circuits for sensing were fabricated on Au/PI bilayer film with a laser cutter, while the remaining parts received a parylene coating. The PEDOT:PSS/PVA hydrogel pH sensors were adhered onto the side of Au/PI film with conductive electrodes, while the drug-release patches were integrated onto the other side. Then, two pieces of e-skin layer were adhered to a piece of PNIPAM hydrogel at the edges, as shown in Fig.  S37A .

Quantification of RhB release from the drug patch: The drug-release patches of PLGA/RhB mixture were immersed into phosphate-buffered saline (~PBS) solutions (pH ~ 5) for 1 h under different temperatures ranging from 25 to 45 °C. The UV-vis absorbance spectrometer was used to analyze the amount of the RhB released from the drug-patch.

Measurement of pH sensitivity: The PEDOT:PSS hybrid hydrogels were submerged into PBS solutions with different pH values ranging from 3 to 8. In 5 min, hydrogels were removed from PBS solutions and their resistance was measured with PowerLab. The effect of pH on the electrical properties of the PEDOT:PSS/PVA hydrogel is attributed to the ionic interaction between PEDOT and PSS polymer chains. Under acidic conditions, PEDOT chains are uniformly distributed along the PSS polymer chains, ensuring the formation of continuous electrical connections between PEDOT segments. While pH shifts from acidic to more alkaline, the homogenous distribution of PEDOT along the PSS polymer chains is interrupted by negatively charged hydroxy groups, and buried inside the insulating PSS phase, as illustrated in Fig.  S37D .

Cell morphology analysis, cell viability test, and histological analysis

Cell morphology analysis: Swiss 3T3-J-2 cells are seeded in a 96-well plate with 10,000 cells per well and cultured for 48 h at 37 °C. Here, the cells are exposed to pure PNIPAM hydrogel, AgNWs/PI nanocomposite, RGO/PI nanocomposite, PEDOT:PSS/PI nanocomposite, MXene/PI nanocomposite, and the integrated soft robot. Notably, each of these functional units was encapsulated with a parylene film (thickness ~ 2 μm). For examining the effect of thermal stimulation on cell viability, the 3T3-J-2 cells with a soft robot are subjected to an environment at 39 °C for a duration of 48 h. To detect Vimentin antigen, chromogenic Immunohistochemistry (IHC) is performed on paraffin-embedded cells that were sectioned at 5 microns. This IHC is carried out using the Leica Bond Rx Autostainer system. Slides are dewaxed in Bond Dewax solution (AR9222) and hydrated in Bond Wash solution (AR9590). Heat-induced antigen retrieval is performed at 100 °C in Bond-Epitope Retrieval solution 1 pH-6.0 (AR9961). After pretreatment, slides are incubated with Vimentin Antibody (5741, Cell Signaling Technologies) at 1:1000 for 30 m followed with Novolink Polymer (RE7260-CE) secondary. Antibody detection with 3,3’-diaminobenzidine (DAB) is performed using the Bond Intense R detection system (DS9263). Stained slides are dehydrated and coverslipped with Cytoseal 60 (8310-4, Thermo Fisher Scientific). A positive control tissue is included for this run. High-resolution acquisition of IF slides is performed with the Aperio Versa 200 scanner (Leica Biosystems Inc.) at an apparent magnification of 20X. Immunofluorescence reaction (IF) is performed on paraffin-embedded cells that are sectioned at 5 microns. This assay is carried out on the Bond Rx fully automated slide staining system (Leica Biosystems) using the Bond Research Detection kit (DS9455). Slides are dewaxed in Bond Dewax solution (AR9222) and hydrated in Bond Wash solution (AR9590). Heat induced antigen retrieval is performed at 100 °C in Bond-Epitope Retrieval solution 1 pH-6.0 (AR9961) for 30 min. After pretreatment, slides are incubated with Vimentin Antibody (5741, Cell Signaling Technologies) at 1:1000 for 30 m. Ready-to use secondary antibody, Leica’s Novolink Polymer (RE7260-CE) is used followed by TSA Cy5 (SAT705A001EA, Akoya Biosciences) to visualize the target of interest. Nuclei were stained with Hoechst 33258 (Invitrogen). The stained slides are mounted with ProLong Gold antifade reagent (P36930, Life Technologies). Positive controls are included for each assay. High resolution acquisition of IF slides is performed with the Aperio Versa 200 scanner (Leica Biosystems Inc.) at an apparent magnification of 20X.

Cell viability test: Swiss 3T3-J-2 cells are seeded in a 96-well plate with 10,000 cells per well and cultured for 48 h. Here, the cells are exposed to pure PNIPAM hydrogel, and the integrated soft robot. For examining the effect of thermal stimulation on cell viability, the 3T3-J-2 cells with a soft robot are subjected to an environment at 39 °C for a duration of 48 h. We use the Aperio Cytoplasmic version 2 algorithm on image regions that were annotated to exclude artifacts. The algorithm input parameters included Clear Area Intensity, Optical Density values (RGB) for both counterstain and biomarker detection, intensity threshold values, and minimum/maximum size and smoothing values for cell segmentation. Several of these parameters were adjusted to apply to the specific marker (Vimentin) and cells.

Histological analysis: Explanted organs were bisected and stored in 10% buffered formalin inside 50 mL conical tubes. Following this, these tissue samples were subsequently prepared for histological evaluation with H&E staining and were imaged using Leica Biosystems.

The fabrication of a robotic thera-gripper: The e-skin layer of the soft robotic gripper is formed using the laser cutting machine, consisting of four strain sensors made of serpentine Au/PI resistors, two Au E-stim electrodes, and two temperature sensors made of thermal resistors, as shown in Fig.  S47A, B . Following the application of a parylene film, four pieces of PNIPAM hydrogel were bonded onto the e-skin layer to form the actuators. The soft robotic gripper can gently hold a mouse heart during its beating movement.

In vivo animal experiment: Procedures used in the study were reviewed and approved by the Institutional Animal Care and Use Committee and Research Animal Resources at the North Carolina University Chapel Hill (IACUC ID:21-241.0). Female mice (weight 20–30 g; age, 10 weeks) were purchased from the Jackson Laboratory. The detailed surgery process is described in Supplementary Note  S7 . The electrocardiography (ECG) and heart function were monitored simultaneously using commercial equipment (PowerLab). The myocardial infarction surgery was conducted by permanently occluding the left coronary artery (LCA). Echocardiography was performed to evaluate cardiac function and histological analysis was performed to assess inflammatory effect of the implantable robotic thera-gripper.

Data availability

All data needed to evaluate the conclusions in the manuscript are present in the manuscript and/or the Supplementary Information. The source data is provided with this manuscript.  Source data are provided with this paper.

Code availability

The custom codes are available. https://doi.org/10.5281/zenodo.11095018 .

Xiao, X., Xiao, X., Lan, Y. & Chen, J. Learning from nature for healthcare, energy, and environment. Innovation 2 , 100135 (2021).

PubMed   PubMed Central   Google Scholar  

Wolf, C. & Linden, D. E. J. Biological pathways to adaptability—interactions between genome, epigenome, nervous system and environment for adaptive behavior. Genes Brain Behav. 11 , 3–28 (2012).

Article   CAS   PubMed   Google Scholar  

Ren, L. et al. Biology and bioinspiration of soft robotics: actuation, sensing, and system integration. IScience 24 , 103075 (2021).

Article   PubMed   PubMed Central   ADS   Google Scholar  

Son, D. et al. Multifunctional wearable devices for diagnosis and therapy of movement disorders. Nat. Nanotechnol. 9 , 397–404 (2014).

Article   CAS   PubMed   ADS   Google Scholar  

Choi, H. et al. Adhesive bioelectronics for sutureless epicardial interfacing. Nat. Electron. 6 , 779–789 (2023).

Article   Google Scholar  

Johansson, R. S. & Flanagan, J. R. Coding and use of tactile signals from the fingertips in object manipulation tasks. Nat. Rev. Neurosci. 10 , 345–359 (2009).

Bartolozzi, C., Natale, L., Nori, F. & Metta, G. Robots with a sense of touch. Nat. Mater. 15 , 921–925 (2016).

Chortos, A., Liu, J. & Bao, Z. Pursuing prosthetic electronic skin. Nat. Mater. 15 , 937–950 (2016).

Heng, W., Solomon, S. & Gao, W. Flexible electronics and devices as human–machine interfaces for medical robotics. Adv. Mater. 34 , 2107902 (2022).

Article   CAS   Google Scholar  

Yu, Y. et al. All-printed soft human-machine interface for robotic physicochemical sensing. Sci. Robot. 7 , eabn0495 (2023).

Cianchetti, M., Laschi, C., Menciassi, A. & Dario, P. Biomedical applications of soft robotics. Nat. Rev. Mater. 3 , 143–153 (2018).

Article   ADS   Google Scholar  

Zhao, Q., Li, C., Shum, H. C. & Du, X. Shape-adaptable biodevices for wearable and implantable applications. Lab Chip 20 , 4321–4341 (2020).

Dupont, P. E. et al. A decade retrospective of medical robotics research from 2010 to 2020. Sci. Robot. 6 . https://doi.org/10.1126/scirobotics.abi8017 (2021).

Payne, C. J. et al. An implantable extracardiac soft robotic device for the failing heart: mechanical coupling and synchronization. Soft Robot 4 , 241–250 (2017).

Article   PubMed   Google Scholar  

Rivkin, B. et al. Shape‐controlled flexible microelectronics facilitated by integrated sensors and conductive polymer actuators. Adv. Intell. Syst. 3 , 2000238 (2021).

Zheng, Z. et al. Ionic shape-morphing microrobotic end-effectors for environmentally adaptive targeting, releasing, and sampling. Nat. Commun . 12 . https://doi.org/10.1038/s41467-020-20697-w (2021).

Banerjee, H., Suhail, M. & Ren, H. Hydrogel actuators and sensors for biomedical soft robots: Brief overview with impending challenges. Biomimetics 3 , 1–40 (2018).

Tse, Z. T. H. et al. Soft robotics in medical applications. J. Med. Robot. Res . 3 . https://doi.org/10.1142/S2424905X18410064 (2018).

Hu, L. et al. An implantable soft robotic ventilator augments inspiration in a pig model of respiratory insufficiency. Nat. Biomed. Eng . https://doi.org/10.1038/s41551-022-00971-6 (2022).

Arab Hassani, F., Jin, H., Yokota, T., Someya, T. & Thakor, N. V. Soft sensors for a sensing-actuation system with high bladder voiding efficiency. Sci. Adv. 6 , 2–10 (2020).

Roh, Y. et al. Vital signal sensing and manipulation of a microscale organ with a multifunctional soft gripper. Sci. Robot. 6 , 1–12 (2021).

Ghosh, A. et al. Gastrointestinal-resident, shape-changing microdevices extend drug release in vivo. Sci. Adv . 6 . https://doi.org/10.1126/sciadv.abb4133 (2020).

Rosalia, L. et al. A soft robotic sleeve mimicking the haemodynamics and biomechanics of left ventricular pressure overload and aortic stenosis. Nat. Biomed. Eng. 6 , 1134–1147 (2022).

Article   PubMed   PubMed Central   Google Scholar  

Wu, Z. et al. A microrobotic system guided by photoacoustic computed tomography for targeted navigation in intestines in vivo. Sci. Robot. 4 , eaax0613 (2019).

Roche, E. T. et al. Soft robotic sleeve supports heart function. Sci. Transl. Med. 9 , 1–12 (2017).

Capella, V. et al. Cytotoxicity and bioadhesive properties of poly-N-isopropylacrylamide hydrogel. Heliyon 5 , e01474 (2019).

Article   CAS   PubMed   PubMed Central   Google Scholar  

Haq, M. A., Su, Y. & Wang, D. Mechanical properties of PNIPAM based hydrogels: a review. Mater. Sci. Eng. C. 70 , 842–855 (2017).

Xu, J., & Song, J. Thermal Responsive Shape Memory Polymers for Biomedical Applications (ed Fazel-Rezai, R.) Ch. 6 (IntechOpen, Rijeka, 2011). https://doi.org/10.5772/19256 .

Czerner, M., Fellay, L. S., Suárez, M. P., Frontini, P. M. & Fasce, L. A. Determination of elastic modulus of gelatin gels by indentation experiments. Procedia Mater. Sci. 8 , 287–296 (2015).

Reeder, J. et al. Mechanically adaptive organic transistors for implantable electronics. Adv. Mater. 26 , 4967–4973 (2014).

Liu, H. et al. 3D printed flexible strain sensors: from printing to devices and signals. Adv. Mater. 33 , 2004782 (2021).

Zhu, Z., Park, H. S. & McAlpine, M. C. 3D printed deformable sensors. Sci. Adv. 6 , eaba5575 (2023).

Qiu, Y. et al. Nondestructive identification of softness via bioinspired multisensory electronic skins integrated on a robotic hand. Npj Flex. Electron . 6 . https://doi.org/10.1038/s41528-022-00181-9 (2022).

Lee, Y. et al. Mimicking human and biological skins for multifunctional skin electronics. Adv. Funct. Mater . 30 . https://doi.org/10.1002/adfm.201904523 (2020).

Afanasenkau, D. et al. Rapid prototyping of soft bioelectronic implants for use as neuromuscular interfaces. Nat. Biomed. Eng. 4 , 1010–1022 (2020).

Subbulakshmi Radhakrishnan, S., Sebastian, A., Oberoi, A., Das, S. & Das, S. A biomimetic neural encoder for spiking neural network. Nat. Commun. 12 , 1–10 (2021).

Li, F. et al. A skin-inspired artificial mechanoreceptor for tactile enhancement and integration. ACS Nano 15 , 16422–16431 (2021).

Lim, C. et al. Tissue-like skin-device interface for wearable bioelectronics by using ultrasoft, mass-permeable, and low-impedance hydrogels. Sci. Adv. 7 , 1–12 (2021).

Minev, I. R. et al. Electronic dura mater for long-term multimodal neural interfaces. Science 347 , 159–163 (2015).

Lee, D. S. et al. Comparison of in vivo biocompatibilities between parylene-C and polydimethylsiloxane for implantable microelectronic devices. Bull. Mater. Sci. 36 , 1127–1132 (2013).

Zhao, Y. et al. Somatosensory actuator based on stretchable conductive photothermally responsive hydrogel. Sci. Robot. 6 , eabd5483 (2021).

Matsumoto, K., Sakikawa, N. & Miyata, T. Thermo-responsive gels that absorb moisture and ooze water. Nat. Commun. 9 , 1–7 (2018).

Article   CAS   ADS   Google Scholar  

Xu, X. et al. Poly(N-isopropylacrylamide)-based thermoresponsive composite hydrogels for biomedical applications. Polymers 12 . https://doi.org/10.3390/polym12030580 (2020).

Roy, D., Brooks, W. L. A. & Sumerlin, B. S. New directions in thermoresponsive polymers. Chem. Soc. Rev. 42 , 7214–7243 (2013).

Tang, L. et al. Poly(N-isopropylacrylamide)-based smart hydrogels: design, properties and applications. Prog. Mater. Sci . 115 . https://doi.org/10.1016/j.pmatsci.2020.100702 (2021).

Ilami, M., Bagheri, H., Ahmed, R., Skowronek, E. O. & Marvi, H. Materials, actuators, and sensors for soft bioinspired robots. Adv. Mater. 33 , 1–47 (2021).

Shojaeifard, M., Niroumandi, S. & Baghani, M. Programming shape-shifting of flat bilayers composed of tough hydrogels under transient swelling. Acta Mech 233 , 213–232 (2022).

Shian, S., Bertoldi, K. & Clarke, D. R. Dielectric elastomer based “grippers” for soft robotics. Adv. Mater. 27 , 6814–6819 (2015).

Basarir, F., Madani, Z., & Vapaavuori, J. Recent advances in silver nanowire based flexible capacitive pressure sensors: from structure, fabrication to emerging applications. Adv. Mater . Interfaces 9 . https://doi.org/10.1002/admi.202200866 (2022).

Amara, U., Hussain, I., Ahmad, M., Mahmood, K. & Zhang, K. 2D MXene-based biosensing: a review. Small 19 , 1–38 (2023).

Chauhan, N., Maekawa, T. & Kumar, D. N. S. Graphene based biosensors—accelerating medical diagnostics to new-dimensions. J. Mater. Res. 32 , 2860–2882 (2017).

Driscoll, N. et al. MXene-infused bioelectronic interfaces for multiscale electrophysiology and stimulation. Sci. Transl. Med. 13 , eabf8629 (2021).

Shin, Y.-E. et al. Ultrasensitive multimodal tactile sensors with skin-inspired microstructures through localized ferroelectric polarization. Adv. Sci. 9 , 2105423 (2022).

Hua, Q. et al. Skin-inspired highly stretchable and conformable matrix networks for multifunctional sensing. Nat. Commun. 9 , 244 (2018).

Lv, P., Cheng, H., Ji, C., & Wei, W. Graphitized-rGO/polyimide aerogel as the compressible thermal interface material with both high in-plane and through-plane thermal conductivities. Materials 14 . https://doi.org/10.3390/ma14092350 (2021).

Du, F. P. et al. PEDOT:PSS/graphene quantum dots films with enhanced thermoelectric properties via strong interfacial interaction and phase separation. Sci. Rep. 8 , 1–12 (2018).

Zhang, W. et al. High-efficiency ITO-free polymer solar cells using highly conductive PEDOT:PSS/surfactant bilayer transparent anodes. Energy Environ. Sci. 6 , 1956–1964 (2013).

Kim, S. H. et al. Nanoscale chemical and electrical stabilities of graphene-covered silver nanowire networks for transparent conducting electrodes. Sci. Rep. 6 , 33074 (2016).

Article   CAS   PubMed   PubMed Central   ADS   Google Scholar  

Zhang, T., Zhao, Y. & Wang, K. Polyimide aerogels cross-linked with aminated ag nanowires: mechanically strong and tough. Polymers (Basel) 9 , 530 (2017).

Fang, J. et al. A shift from efficiency to adaptability: recent progress in biomimetic interactive soft robotics in wet environments. Adv. Sci. 9 , 2104347 (2022).

Zhu, J. et al. Intelligent soft surgical robots for next-generation minimally invasive surgery. Adv. Intell. Syst. 3 , 2100011 (2021).

Sezer Hicyilmaz, A. Applications of polyimide coatings: a review. SN Appl. Sci. 3 , 363 (2021).

Zhang, L. & Feng, G. A one-step-assembled three-dimensional network of silver/polyvinylpyrrolidone (PVP) nanowires and its application in energy storage. Nanoscale 12 . https://doi.org/10.1039/d0nr00991a (2020).

Zhang, L., Liu, X., Deb, A., & Feng, G. Ice-templating synthesis of hierarchical and anisotropic silver-nanowire-fabric aerogel and its application for enhancing thermal energy storage composites. ACS Sustain . Chem. Eng . 7 . https://doi.org/10.1021/acssuschemeng.9b05413 (2019).

Chang, B. et al. Reduced graphene oxide promoted assembly of graphene@polyimide film as a flexible cathode for high-performance lithium-ion battery. RSC Adv 10 , 8729–8734 (2020).

Dhakshnamoorthy, M., Vikram, S. & Vasanthakumari, R. Development of flexible low dielectric constant polyimide films based on iso-propylidene, aryl-ether linked dianhydride/diamine. Int. J. Sci. Eng. Res. 3 , 5 (2012).

Google Scholar  

Peng, X. et al. Thermoresponsive deformable actuators prepared by local electrochemical reduction of poly(n-isopropylacrylamide)/graphene oxide hydrogels. ACS Appl. Nano Mater. 1 , 1522–1530 (2018).

Spratte, T. et al. Thermoresponsive hydrogels with improved actuation function by interconnected microchannels. Adv. Intell. Syst. 4 , 2100081 (2022).

Fan, W. et al. Precisely controlling the output force of hydrogel actuator based on thermodynamic nonequilibrium temporary deformation. ACS Appl. Mater. Interfaces. 12 , 49042–49049 (2020).

Wei, S. et al. Bioinspired synergistic fluorescence-color-switchable polymeric hydrogel actuators. Angew. Chemie Int. Ed. 58 , 16243–16251 (2019).

Gao, G. et al. Snap-buckling motivated controllable jumping of thermo-responsive hydrogel bilayers. ACS Appl. Mater. Interfaces. 10 , 41724–41731 (2018).

Zheng, S. Y. et al. Programmed deformations of 3D‐printed tough physical hydrogels with high response speed and large output force. Adv. Funct. Mater . 28 , 1803366 (2018).

Cai, G., Ciou, J.-H., Liu, Y., Jiang, Y. & Lee, P. S. Leaf-inspired multi responsive MXene-based actuator for programmable smart devices. Sci. Adv. 5 , eaaw7956 (2023).

Li, J. et al. Highly bidirectional bendable actuator engineered by LCST–UCST bilayer hydrogel with enhanced interface. ACS Appl. Mater. Interfaces. 12 , 55290–55298 (2020).

Garg, R. & Vitale, F. Latest advances on MXenes in biomedical research and health care. MRS Bull. 48 , 283–290 (2023).

Tang, Y. et al. Wireless miniature magnetic phase-change soft actuators. Adv. Mater. 34 , 2204185 (2022).

Kim, Y., Parada, G. A., Liu, S. & Zhao, X. Ferromagnetic soft continuum robots. Sci. Robot. 4 , eaax7329 (2019).

van Rhoon, G. C. et al. CEM43 °C thermal dose thresholds: a potential guide for magnetic resonance radiofrequency exposure levels? Eur. Radiol. 23 , 2215–2227 (2013).

Dewhirst, M. W., Viglianti, B. L., Lora-Michiels, M., Hoopes, P. J. & Hanson, M. Thermal dose requirement for tissue effect: experimental and clinical findings. Proc. SPIE Int. Soc. Opt. Eng. 4954 , 37 (2003).

PubMed   PubMed Central   ADS   Google Scholar  

Won, S. M., Cai, L., Gutruf, P., & Rogers, J. A. Wireless and battery-free technologies for neuroengineering. Nat. Biomed. Eng . https://doi.org/10.1038/s41551-021-00683-3 (2021).

Koo, J. et al. Wirelessly controlled, bioresorbable drug delivery device with active valves that exploit electrochemically triggered crevice corrosion. Sci. Adv. 6 , eabb1093 (2023).

Zhou, H., Mayorga-Martinez, C. C., Pané, S., Zhang, L. & Pumera, M. Magnetically driven micro and nanorobots. Chem. Rev. 121 , 4999–5041 (2021).

Hannan, M. A., Mutashar, S., Samad, S. A. & Hussain, A. Energy harvesting for the implantable biomedical devices: issues and challenges. Biomed. Eng. Online. 13 , 79 (2014).

Kanaan, A. I., & Sabaawi, A. M. A. Implantable Wireless Systems: A Review of Potentials and Challenges (eds Al-Rizzo, H. & Abushamleh S.) (IntechOpen, Rijeka, 2021). https://doi.org/10.5772/intechopen.99064 .

Lu, D. et al. Bioresorbable wireless sensors as temporary implants for in vivo measurements of pressure. Adv. Funct. Mater. 30 , 2003754 (2020).

Lu, D. et al. Bioresorbable, wireless, passive sensors as temporary implants for monitoring regional body temperature. Adv. Healthc. Mater. 9 , 2000942 (2020).

Jeong, Y. R. et al. A skin-attachable, stretchable integrated system based on liquid GaInSn for wireless human motion monitoring with multi-site sensing capabilities. NPG Asia Mater. 9 , e443–e443 (2017).

Keum, D. H. et al. Wireless smart contact lens for diabetic diagnosis and therapy. Sci. Adv. 6 , eaba3252 (2023).

Traficante, D. D. Impedance: What it is, and why it must be matched. Concepts Magn. Reson. 1 , 73–92 (1989).

Abdolkhani, A. Fundamentals of Inductively Coupled Wireless Power Transfer Systems (ed Coca, E.) (IntechOpen, Rijeka, 2016). https://doi.org/10.5772/63013 .

Kiani, M. & Ghovanloo, M. The circuit theory behind coupled-mode magnetic resonance-based wireless power transmission. IEEE Trans. Circuits Syst. I Regul. Pap. 59 , 2065–2074 (2012).

Article   MathSciNet   PubMed   PubMed Central   Google Scholar  

Das Barman, S., Reza, A. W., Kumar, N., Karim, M. E. & Munir, A. B. Wireless powering by magnetic resonant coupling: recent trends in wireless power transfer system and its applications. Renew. Sustain. Energy Rev. 51 , 1525–1552 (2015).

Imura, T. Wireless power transfer: using magnetic and electric resonance coupling techniques. https://doi.org/10.1007/978-981-15-4580-1 2020.

Mozzini, C., Xotta, G., Garbin, U., Pasini, A. M. F. & Cominacini, L. Non-exertional heatstroke: a case report and review of the literature. Am. J. Case Rep. 18 , 1058–1065 (2017).

Hsiao, J.-H., Chang, J. Y. & Cheng, C.-M. Soft medical robotics: clinical and biomedical applications, challenges, and future directions. Adv. Robot. 33 , 1099–1111 (2019).

Laschi, C., Mazzolai, B. & Cianchetti, M. Soft robotics: technologies and systems pushing the boundaries of robot abilities. Sci. Robot. 1 , eaah3690 (2016).

Ochodnicky, P., Cruz, C. D., Yoshimura, N. & Cruz, F. Neurotrophins as regulators of urinary bladder function. Nat. Rev. Urol. 9 , 628–637 (2012).

Tudor, K. I., Sakakibara, R. & Panicker, J. N. Neurogenic lower urinary tract dysfunction: evaluation and management. J. Neurol. 263 , 2555–2564 (2016).

Ji, H. et al. Biocompatible in situ polymerization of multipurpose polyacrylamide-based hydrogels on skin via silver ion catalyzation. ACS Appl. Mater. Interfaces. 12 , 31079–31089 (2020).

Grill, W.M. Electrical stimulation for control of bladder function. Proc. 31st Annual International Conference of the IEEE Engineering in Medicine and Biology Society Engineering Future Biomedical 2369–2370 (EMBC, 2009). https://doi.org/10.1109/IEMBS.2009.5335001 .

Coolen, R. L., Groen, J. & Blok, B. F. M. Electrical stimulation in the treatment of bladder dysfunction: technology update. Med. Devices Evid. Res. 12 , 337–345 (2019).

Zhang, Y. et al. Climbing-inspired twining electrodes using shape memory for peripheral nerve stimulation and recording. Sci. Adv . 5 . https://doi.org/10.1126/sciadv.aaw1066 (2019).

Zheng, H. et al. A shape-memory and spiral light-emitting device for precise multisite stimulation of nerve bundles. Nat. Commun. 10 , 1–14 (2019).

Song, S., Fallegger, F., Trouillet, A., Kim, K. & Lacour, S. P. Deployment of an electrocorticography system with a soft robotic actuator. Sci. Robot. 8 , eadd1002 (2023).

Woodington, B. J. et al. Electronics with shape actuation for minimally invasive spinal cord stimulation. Sci. Adv. 7 , 1–9 (2021).

Whyte, W. et al. Sustained release of targeted cardiac therapy with a replenishable implanted epicardial reservoir /692/4019/2773 /639/301/54/152 /14/5 /14/35 /14/63 /59/5 /96/106 /96/100 /96/34 article. Nat. Biomed. Eng. 2 , 416–428 (2018).

Chen, J. C. et al. A wireless millimetric magnetoelectric implant for the endovascular stimulation of peripheral nerves. Nat. Biomed. Eng. 6 , 706–716 (2022).

Liu, J. et al. Syringe-injectable electronics. Nat. Nanotechnol. 10 , 629–635 (2015).

Guo, Z. et al. A flexible neural implant with ultrathin substrate for low-invasive brain–computer interface applications. Microsystems Nanoeng. 8 , 1–12 (2022).

Montgomery, M. et al. Flexible shape-memory scaffold for minimally invasive delivery of functional tissues. Nat. Mater. 16 , 1038–1046 (2017).

Xie, C. et al. Three-dimensional macroporous nanoelectronic networks as minimally invasive brain probes. Nat. Mater. 14 , 1286–1292 (2015).

Jiang, Y. et al. Wireless, closed-loop, smart bandage with integrated sensors and stimulators for advanced wound care and accelerated healing. Nat. Biotechnol. 41 , 652–662 (2023).

Sim, K. et al. An epicardial bioelectronic patch made from soft rubbery materials and capable of spatiotemporal mapping of electrophysiological activity. Nat. Electron. 3 , 775–784 (2020).

Choi, Y. S. et al. Fully implantable and bioresorbable cardiac pacemakers without leads or batteries. Nat. Biotechnol. 39 , 1228–1238 (2021).

Baddour, L. M. et al. Update on cardiovascular implantable electronic device infections and their management: a scientific statement from the American Heart Association. Circulation 121 , 458–477 (2010).

Liu, Y. et al. Soft and elastic hydrogel-based microelectronics for localized low-voltage neuromodulation. Nat. Biomed. Eng. 3 , 58–68 (2019).

Choi, S. et al. Highly conductive, stretchable and biocompatible Ag–Au core–sheath nanowire composite for wearable and implantable bioelectronics. Nat. Nanotechnol. 13 , 1048–1056 (2018).

Monteiro, L. M., Vasques-Nóvoa, F., Ferreira, L., Pinto-Do-ó, P. & Nascimento, D. S. Restoring heart function and electrical integrity: Closing the circuit. Npj Regen. Med. 2 , 1–13 (2017).

ADS   Google Scholar  

Cao, H., Kang, B. J., Lee, C. A., Shung, K. K. & Hsiai, T. K. Electrical and mechanical strategies to enable cardiac repair and regeneration. IEEE Rev. Biomed. Eng. 8 , 114–124 (2015).

Tsao, C. W. et al. Heart disease and stroke statistics—2022 update: a report from the American Heart Association. Circulation 145 , e153–e639 (2022).

Jafari Tadi, M. et al. Gyrocardiography: a new non-invasive monitoring method for the assessment of cardiac mechanics and the estimation of hemodynamic variables. Sci. Rep. 7 , 6823 (2017).

Hwang, J. C. et al. In situ diagnosis and simultaneous treatment of cardiac diseases using a single-device platform. Sci. Adv. 8 , eabq0897 (2023).

Recco, D. P., Roy, N., Gregory, A. J. & Lobdell, K. W. Invasive and noninvasive cardiovascular monitoring options for cardiac surgery. JTCVS Open 10 , 256–263 (2022).

Lindsey, M. L., Kassiri, Z., Virag, J. A. I., de Castro Brás, L. E. & Scherrer-Crosbie, M. Guidelines for measuring cardiac physiology in mice. Am. J. Physiol. Circ. Physiol. 314 , H733–H752 (2018).

Chengode, S. Left ventricular global systolic function assessment by echocardiography. Ann. Card. Anaesth. 19 , 26 (2016).

Nishimura, R. A. et al. Doppler echocardiography: theory, instrumentation, technique, and application. Mayo Clin. Proc. 60 , 321–343 (1985).

Article   MathSciNet   CAS   PubMed   Google Scholar  

Download references

Acknowledgements

This work was supported by the start-up funds from University of North Carolina at Chapel Hill and the fund from National Science Foundation (award # ECCS-2139659). We thank Yongjuan Xia in the Pathology Services Core (PSC) for expert technical assistance with Histopathology/Digital Pathology including tissue sectioning, immunohistochemical staining, and imaging. The PSC is supported in part by an NCI Center Core Support Grant (P30CA016086). We thank Dr. Amar S. Kumbhar in the Chemistry Department at UNC Chapel Hill for their help with SEM imaging. Research reported in this publication was also supported by the National Institute of Biomedical Imaging and Bioengineering at the National Institutes of Health under award number 1R01EB034332-01. This work was performed in part at the Chapel Hill Analytical and Nanofabrication Laboratory, CHANL, a member of the North Carolina Research Triangle Nanotechnology Network, RTNN, which is supported by the National Science Foundation, Grant ECCS-2025064, as part of the National Nanotechnology Coordinated Infrastructure, NNCI.

Author information

Authors and affiliations.

Department of Applied Physical Sciences, University of North Carolina, Chapel Hill, NC, 27514, USA

Lin Zhang, Yihang Wang, Yihan Liu, Yizhang Wu, Wanrong Xie, Chuqi Huang & Wubin Bai

Department of Biomedical Engineering, University of North Carolina, Chapel Hill, NC, 27514, USA

Sicheng Xing, Hiep Thanh Tran & Wei Luo

MCAllister Heart Institute Core, University of North Carolina, Chapel Hill, NC, 27514, USA

Haifeng Yin

Department of Biology, University of North Carolina, Chapel Hill, NC, 27514, USA

Hannah Weisbecker, Amber Huang, Brynn Brusseau & Yici Zou

Department of Chemistry, University of North Carolina, Chapel Hill, NC, 27514, USA

Ziheng Guo, Michael Demaesschalck, Collin McKinney & Samuel Hankley

Joint Department of Biomedical Engineering, North Carolina State University, Raleigh, NC, 27606, USA

Tianhong Han

Weldon School of Biomedical Engineering, Purdue University, West Lafayette, IN, 47907, USA

Jett Messenger

You can also search for this author in PubMed   Google Scholar

Contributions

L.Z. and W.B. conceived the ideas and designed the research. L.Z., S.X., H.W., H.T., Z.G., T.H., C.H., Y.Z.W., Y.L., W.L., M.D., S.H., A.H., and B.B. fabricated and characterized the samples. L.Z. performed the mechanical modeling and simulation. L.Z., H.W., H.T., Z.G., and H.Y. performed the ex-vivo and in-vivo studies. L.Z., S.X., H.W., Y.L., Z.G., Y.H.W., W.X., Y.Z., J.M., H.T., and C.M. performed the data analysis. L.Z. and W.B. wrote the manuscript with input from all authors.

Corresponding author

Correspondence to Wubin Bai .

Ethics declarations

Competing interests.

The University of North Carolina at Chapel Hill (No. 035052/602726: filed on Oct. 13, 2023) filed a provisional patent application, titled “Skin-inspired, sensory robots for electronic implants”, surrounding this work, where W.B. and L.Z. are the co-inventors. The remaining authors declare no competing interests.

Peer review

Peer review information.

Nature Communications thanks Donghee Son, and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. A peer review file is available.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Supplementary information, supplementary movie 1, supplementary movie 2, supplementary movie 3, supplementary movie 4, supplementary movie 5, supplementary movie 6, editorial policy checklist, peer review file, source data, rights and permissions.

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ .

Reprints and permissions

About this article

Cite this article.

Zhang, L., Xing, S., Yin, H. et al. Skin-inspired, sensory robots for electronic implants. Nat Commun 15 , 4777 (2024). https://doi.org/10.1038/s41467-024-48903-z

Download citation

Received : 13 December 2023

Accepted : 15 May 2024

Published : 05 June 2024

DOI : https://doi.org/10.1038/s41467-024-48903-z

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

By submitting a comment you agree to abide by our Terms and Community Guidelines . If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.

Quick links

• Explore articles by subject
• Guide to authors
• Editorial policies

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

• Corpus ID: 270559764

Optical Variability Properties of Southern TESS Blazars

• Ryne Dingler , Krista Lynne Smith
• Published 14 June 2024

Ask This Paper

By using this feature, you agree to AI2's terms and conditions and that you will not submit any sensitive or confidential info.

AI2 may include your prompts and inputs in a public dataset for future AI research and development. Please check the box to opt-out.

Ask a question about " "

Supporting statements, figures and tables from this paper.

12 References

Tess-point: high precision tess pointing tool, astrocut: tools for creating cutouts of tess images, optimal data-based binning for histograms and histogram-based probability density models, statistical challenges in modern astronomy ii, related papers.

Showing 1 through 3 of 0 Related Papers

An official website of the United States government Here’s how you know keyboard_arrow_down

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Jump to main content

Patent Public Search

The Patent Public Search tool is a new web-based patent search application that will replace internal legacy search tools PubEAST and PubWEST and external legacy search tools PatFT and AppFT. Patent Public Search has two user-selectable modern interfaces that provide enhanced access to prior art. The new, powerful, and flexible capabilities of the application will improve the overall patent searching process.

If you are  new to patent searches , or want to use the  functionality that was available in the USPTO’s PatFT/AppFT , select Basic search to look for patents by keywords or common fields, such as inventor or publication number. Select Advanced search to use full query options as well as to further filter a patent search by database or organize documents through tagging.

Basic search            Advanced search

This is the new Patent Public Search page. The URL has changed. Please update your bookmarks or favorites.

Search references.

• Searchable indexes
• Keyboard shortcuts
• Case sensitivity

Training materials

• Quick reference guides
• Tutorial videos

Support & Assistance

• Frequently asked questions

Additional information about this page

IEEE Account

• Change Username/Password
• Update Address

Purchase Details

• Payment Options
• Order History
• View Purchased Documents

Profile Information

• Communications Preferences
• Profession and Education
• Technical Interests
• US & Canada: +1 800 678 4333
• Worldwide: +1 732 981 0060
• Contact & Support
• About IEEE Xplore
• Accessibility
• Terms of Use
• Nondiscrimination Policy
• Privacy & Opting Out of Cookies

A not-for-profit organization, IEEE is the world's largest technical professional organization dedicated to advancing technology for the benefit of humanity. © Copyright 2024 IEEE - All rights reserved. Use of this web site signifies your agreement to the terms and conditions.

An ultra-miniaturized equipoise swastik-shaped THz absorber

• Published: 20 June 2024
• Volume 56 , article number  1232 , ( 2024 )

Cite this article

• Anbuselvi Mathivanan 1 ,
• Saravanan Palaniswamy 1 &
• Gulam Nabi Alsath Mohammed 2  

A novel ultra-miniaturized metamaterial-based absorber is proposed in this paper for THz shielding applications. A novel Swastik shape-based absorber is designed using a gold conductor with increased electrical length. The focus of this research work is to design an absorber geometry with minimal footprint and maximum absorptivity. As the geometry of the absorber is minimized, the operating frequency shifts to a higher space. However, in this proposed absorber, the optimal design is attained with minimal footprint and low operating frequency. The proposed absorber has an overall footprint of 0.129λeff × 0.129λeff for the operating frequency of 1.783 THz. The metamaterial’s absorption properties are analyzed to examine the proposed THz absorber. In free space, the absorptivity is estimated at 100% at its operating frequency. The goodness of the proposed THz absorber is enhanced by the polarization insensitivity of its rotationally symmetrical geometry. The analysis of the parametric variations is presented, along with the variation in the thickness of the analyte. Thus, the proposed THz metamaterial-based absorber is miniaturized by at most 33% of the existing designs and suitable for different sensing applications.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Price includes VAT (Russian Federation)

Instant access to the full article PDF.

Rent this article via DeepDyve

Institutional subscriptions

Data availability

Not applicable

Ahmed, S., Sungjoon, L.: Biosens. Bioelectron. 117 , 398 (2018)

Article   Google Scholar  

Baqir, M.A., Choudhury, P.K.: IEEE Photonics Techn Let. 35 (4), 183–186 (2023)

Article   ADS   Google Scholar  

Bilal, R.M.H., Naveed, M.A., Baqir, M.A., Ali, M.M., Rahim, A.A.: Opt. Mater. Express. 10 , 3007–3020 (2020)

Butler, L., Wilbert, D.S., Baughman, W., Balci, S., Kung, P., Kim, S.M.: Inter. Semi. Dev. Res. Symp. (ISDRS) 1 (2011)

Cai, S., Pan, H., González-Vila, Á., Guo, T., Gillan, D., Wattiez, R., Caucheteur, C.: Opt. Exp. 28 , 19740 (2020)

Chen, M.I., Singh, L., Xu, N., Singh, R., Zhang, W., Xie, L.: Opt. Exp. 25 , 14089 (2017)

Chen, F., Cheng, Y., Luo, H.: IEEE Access. 8 , 82981 (2020)

Cong, L., Singh, R.: arXiv 1408.3711 (2014)

Fei, Y., Li, L., Wang, R., Tian, H., Liu, J., Jianqiang Liu, et al.: J. Lightwave Tech. 37 , 1103 (2019)

Gelmont, B., Parthasarathy, R., Globus, T., Bykhovski, A., Swami, N.: IEEE Sens. J. 8 , 791 (2008)

Harleen, K., Hari Shankar Singh.: Optik, 250168339 (2022)

Ibraheem Al-Naib, Hebestreit, E., Rockstuhl, C., Lederer, F., Christodoulides, D., et al.: Tsuneyuki Ozaki Phys. Rev. lett. 112 183903 (2014)

Jepsen, P.U., Cooke, D.G., Koch, M.: Laser Phot Rev. 5 , 124 (2011)

Jiu, F., Ruan, Zi, F., Meng, R.Z., Zou, S.M., Pan, Sheng, W., Ji: Microwave Optical Technol. Lett. 65 , 1 (2023)

Jornet, J.M., Akyildiz, I.F.: IEEE J. Sel. Areas Comm. 31 , 685 (2013)

Kumar, S., Guo, Z., Singh, R., Wang, Q., Zhang, B., Cheng, S., et al.: J. Lightwave Tech. 39 , 4069 (2021)

Landy, N.I., Sajuyigbe, S., Mock, J.J., Smith, D.R., Padilla, W.J.: Phys. Rev. Lett. 100 , 207402 (2008)

Ma, A., Zhong, R., Wu, Z., Wang, Y., Yang, L., Liang, Z., et al.: Fron. Phys. 441 (2020)

Muhammad, A.N., Rana, M.H.B., Arbab, A.R., Muhammad Abuzar Baqir, and Muhammad Mahmood Ali.: Appl. Opt. 60 , 9160–9166 (2021)

Google Scholar  

Punnag Padhy, P.K., Sahu, R.: Jh Sens. and Actu. B: Chem. 225 , 115 (2015)

Raeena, M.S.: Anveshkumar Nella, Maheswar Rajagopal. Optik. 261 , 169090 (2022)

ADS   Google Scholar  

Reinhard, B., Klemens, M., Schmitt, V., Wollrab: Jens Neu, René Beigang, and Marco Rahm. App Phys. Lett. 100 , 221101 (2012)

Ruan, J.F., Tu, J.Y., Wang, D.L., Tao, Z., Yuan, Y., Sheng Wei Ji: Opt. Mater. 137 , 113604 (2023a)

Ruan, J.-F., Zhu, D.-W., Tao, Z.: Rui-Zhi Zou & Sheng-Min Pan. J. Electromagn. Waves Appl. 37 14 , 1221–1233 (2023b)

Sheta, E.M., Choudhury, P.K., Abdel-Baset, M.A., Ibrahim: Opt. Mater. 133 , 112990 (2022)

Smith, D.R., Padilla, W.J., Vier, D.C., Nemat-Nasser, S.C., Schultz, S.: Phys. Rev. Lett. 84 , 4184 (2020)

Sreekanth, K.V., Mahalakshmi, P., Song Han, D., Vigneswaran, M.S., Mani Rajan, R., et al.: Jha J. of App. Phys. 128 , 173106 (2020)

Tao, H., Bingham, C.M., Strikwerda, A.C., Pilon, D., Shrekenhamer, D., Landy, N.I., et al.: Phys. Rev. B. 78 , 241103 (2008)

Tao, A.S., Saadeldin Hameed, M.F., O, Elkaramany, E.M.A., Obayya, S.S.A: IEEE Sens. J. 19 , 7993 (2019)

Viktor, G.: Sov Phys. Usp. 10 , 509 (1968)

Wang, Z., Han, Y., Xu, N., Chen, L., Li, C., Wu, L., et al.: IEEE Trans. Tera Sci. Tech. 8 , 161 (2018)

Watts, C.M., Shrekenhamer, D., Montoya, J., Lipworth, G., Hunt, J., Sleasman, T., et al.: Nat. Phot. 8 , 605 (2014)

Wilbert, D.S., Hokmabadi, M.P., Kung, P., Kim, S.M.: IEEE Trans. Tera Sci. Tech. 3 , 846 (2013)

Yahiaoui, R., Tan, S., Cong, L., Singh, R., Yan, F., Zhang, W.: J. App Phys. 118 , 083103 (2015)

Yahiaoui, R., Strikwerda, A.C., Jepsen, P.U.: IEEE Sens. J. 16 , 2484 (2016)

Yu, T., Lang, Chen, H.: Photonics. 8 , 164 (2021)

Yu, S.D., Azad, A.K., Hara, J.F., O, Simakov, E.I.: Phys. Rev. B. 82 , 205117 (2010)

Zakir, S., et al.: IEEE Photonics J. 15 (1), 1–8 (2023)

Zi, F., Meng, Z., Tao: Jiu Fu Ruan, Rui Zhi Zou, Sheng Wei Ji. Phys. Lett. A. 445 , 128269 (2022)

Download references

Not applicable.

Author information

Authors and affiliations.

Sri Sivasubramaniya Nadar College of Engineering, Kalavakkam, Tamil Nadu, 603 110, India

Anbuselvi Mathivanan & Saravanan Palaniswamy

College of Engineering Guindy, Anna University, Chennai, Tamil Nadu, 600 025, India

Gulam Nabi Alsath Mohammed

You can also search for this author in PubMed   Google Scholar

Contributions

A. Literature review, Design, analysis, Paper writing. B. Literature review, Design, Paper writing, Formatting. C. Literature review, Design, Paper writing, Review.

Corresponding author

Correspondence to Anbuselvi Mathivanan .

Ethics declarations

The authors certify that this material or similar material has not been and will not be submitted to or published in any other publication before. Furthermore, the authors certify that they have participated sufficiently in the work to take public responsibility for the content, including participation in the concept, design, analysis, writing, or revision of the manuscript.

Ethics approval and consent to participate

Not Applicable.

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.

Reprints and permissions

About this article

Mathivanan, A., Palaniswamy, S. & Mohammed, G.N.A. An ultra-miniaturized equipoise swastik-shaped THz absorber. Opt Quant Electron 56 , 1232 (2024). https://doi.org/10.1007/s11082-024-06878-7

Download citation

Received : 12 September 2023

Accepted : 20 March 2024

Published : 20 June 2024

DOI : https://doi.org/10.1007/s11082-024-06878-7

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• THz absorber
• Single-band
• Metamaterial

Advertisement

• Find a journal
• Publish with us
• Track your research