Flexible electronic brush: Real-time multimodal sensing powered by reservoir computing through whisker dynamics

Area of Science:

• Robotics
• Sensor Technology
• Artificial Intelligence

Background:

• Soft robotic control relies on multimodal sensing of soft body dynamics.
• Mimicking biological whiskers offers a method for digitizing soft body motion.
• Existing whisker-like sensors struggle to effectively detect slip information.

Purpose of the Study:

• To develop a novel sensor system for comprehensive whisker motion monitoring.
• To address the limitation of existing sensors in detecting slip.
• To enable precise control and digitization of soft robotic movements.

Main Methods:

• Development of a multitasking electronic brush (e-brush) using bundled whiskers.
• Integration of four pressure sensors for multimodal data acquisition (motion, speed, force, slip, surface).
• Application of a reservoir computing (RC) algorithm for extracting complex motion parameters, including slip.

Main Results:

• The e-brush demonstrates long-term, low-pressure detection (as low as 50 pascals).
• The RC algorithm successfully extracts multiple motion parameters, crucially including slip detection.
• Proof-of-concept achieved by accurately detecting the motion trajectory of handwriting.

Conclusions:

• The developed e-brush offers a promising solution for advanced whisker-like sensing in soft robotics.
• This technology enhances the ability to monitor and control soft body dynamics with unprecedented detail.
• The system's success in handwriting motion capture highlights its potential for diverse applications in sensing and robotics.