Sensuator: A Hybrid Sensor–Actuator Approach to Soft Robotic Proprioception Using Recurrent Neural Networks

Preechayasomboon, Pornthep

doi:10.3390/act10020030

Cited by 16 publications

(11 citation statements)

References 32 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Experiments showed that this method is feasible for realizing proprioception and is robust to common sensor failures. 262 There is similar research using soft robotic skin to contact the outside and the collected signals are combined with RNN for analysis. Truby et al used deep learning to build a framework for predicting the 3D configuration of a soft robot from the feedback of soft and proprioceptive sensor skin.…”

Section: Stimuli-response Materials (Srms)mentioning

confidence: 99%

See 1 more Smart Citation

Recent advances in biomimetic soft robotics: fabrication approaches, driven strategies and applications

et al. 2022

View full text Add to dashboard Cite

show abstract

Section: Stimuli-response Materials (Srms)mentioning

confidence: 99%

“…Experiments showed that this method is feasible for realizing proprioception and is robust to common sensor failures. 262…”

Section: Applicationsmentioning

confidence: 99%

Recent advances in biomimetic soft robotics: fabrication approaches, driven strategies and applications

et al. 2022

View full text Add to dashboard Cite

show abstract

“…Thuruthel et al embedded soft sensors into the actuator to obtain bending data and used it to train an RNN that can predict the position of the actuator's tip and the force applied by it [17]. Another group used a Bidirectional Long Short-Term Memory (BiLSTM) network, a type of RNN, to estimate the position of a hydraulic soft hybrid sensor-actuator [18].…”

Section: Related Workmentioning

confidence: 99%

Modeling of Soft Pneumatic Actuators with Different Orientation Angles Using Echo State Networks for Irregular Time Series Data

et al. 2022

View full text Add to dashboard Cite

Modeling of soft robotics systems proves to be an extremely difficult task, due to the large deformation of the soft materials used to make such robots. Reliable and accurate models are necessary for the control task of these soft robots. In this paper, a data-driven approach using machine learning is presented to model the kinematics of Soft Pneumatic Actuators (SPAs). An Echo State Network (ESN) architecture is used to predict the SPA’s tip position in 3 axes. Initially, data from actual 3D printed SPAs is obtained to build a training dataset for the network. Irregular-intervals pressure inputs are used to drive the SPA in different actuation sequences. The network is then iteratively trained and optimized. The demonstrated method is shown to successfully model the complex non-linear behavior of the SPA, using only the control input without any feedback sensory data as additional input to the network. In addition, the ability of the network to estimate the kinematics of SPAs with different orientation angles θ is achieved. The ESN is compared to a Long Short-Term Memory (LSTM) network that is trained on the interpolated experimental data. Both networks are then tested on Finite Element Analysis (FEA) data for other θ angle SPAs not included in the training data. This methodology could offer a general approach to modeling SPAs with varying design parameters.

show abstract

“…Xie et al [9,10] proposed a 3D-printed pipe-climbing robot actuated by the flat modular pneumatic artificial muscle consisting of thin-film cylinder modules. Preechayasomboon et al [11] presented a hybrid bellows-style sensor-actuator with salt water inside. Gravagne et al [12] studied a class of cable-driven robot manipulators termed continuum robots such as the Clemson tentacle manipulator, consisting of two independent sections on a continuous backbone with a thin elastic rod.…”

Section: Introductionmentioning

confidence: 99%

A Pneumatic Generator Based on Gas-Liquid Reversible Transition for Soft Robots

et al. 2021

View full text Add to dashboard Cite

The soft robots actuated by pressure, cables, thermal, electrosorption, combustion and smart materials are usually faced with the problems of poor portability, noise, weak load capacity, small deformation and high driving voltages. In this paper, a novel pneumatic generator for soft robots based on the gas-liquid reversible transition is proposed, which has the advantages of large output force, easy deformation, strong load capacity and high flexibility. The pressure of the pneumatic generator surges or drops flexibly through the reversible transformation between liquid and gas phase, making the soft actuator stretch or contract regularly, without external motors, compressors and pressure-regulating components. The gas-liquid reversible-transition actuation process is modeled to analyze its working mechanism and characteristics. The pressure during the pressurization stage increases linearly with a rate regulated by the heating power and gas volume. It decreases exponentially with the exponential term as a quadratic function of time at the fast depressurization stage, while with the exponential term as a linear function of time at the slow depressurization stage. The drop rate can be adjusted by changing the gas volume and cooling conditions. Furthermore, effectiveness has been verified through experiments of the prototype. The pressure reaches 25 bar with a rising rate of +3.935 bar/s when 5 mL weak electrolyte solution is heated at 800 W, and the maximum depressurization rate in air cooling is –3.796 bar/s. The soft finger actuated by the pneumatic generator can bend with an angular displacement of 67.5°. The proposed pneumatic generator shows great potential to be used for the structure, driving and sensing integration of artificial muscles.

show abstract

Sensuator: A Hybrid Sensor–Actuator Approach to Soft Robotic Proprioception Using Recurrent Neural Networks

Cited by 16 publications

References 32 publications

Recent advances in biomimetic soft robotics: fabrication approaches, driven strategies and applications

Recent advances in biomimetic soft robotics: fabrication approaches, driven strategies and applications

Modeling of Soft Pneumatic Actuators with Different Orientation Angles Using Echo State Networks for Irregular Time Series Data

A Pneumatic Generator Based on Gas-Liquid Reversible Transition for Soft Robots

Contact Info

Product

Resources

About