NRI: FND: Scalable, Customizable Sensory Solutions for Dexterous Robotic Hands

点击翻译按钮获取中文摘要

This project aims to enhance the sense of touch for robotic hands. The main goal is to develop prosthetic hands with a sense of touch. The touch sensor's primary application is for upper limb amputees. The research team plans both fundamental research and its application. The types of applications of this kind of sensor include: humanoid robots for assistive work and elder-care, surgical robotics, underwater robotic manipulators and spacesuits. Additionally, educational initiatives including internships, training of under-represented minority, and research experiences and summer modules for high school students are planned through the Johns Hopkins Center for Talented Youth. The technical goal of the project is to build a highly scalable sensor design mimicking different tactile receptors in human skins and encode information from the sensors in a manner analogous to the neural activity of the tactile receptors. The sensor will encode the tactile information at multiple scales, firstly based on different receptor properties, and secondly based on neuron-like encoding of the sensor signals. This technique of encoding the neural activity, known as neuromorphic encoding, converts the sensor activity as an event stream, and from that data obtains finer features. The receptor based sensing along with the various neural network algorithms, for the first time, will provide an approach to texture and shape recognition and, as such, can also be useful for intelligent palpation and tactile perception by dexterous robotic hands.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.