Next-Generation Safe and Dexterous Robot Hands using High-Torque Direct-Drive Actuation

点击翻译按钮获取中文摘要

Skillfully and safely manipulating objects with a sufficiently strong robot hand is highly desirable for factory and warehouse automation applications. However, today’s robot hands cannot simultaneously offer high strength with dexterity and safety due to major limitations in actuation technology. Actuation technology is the system that signals for the robot to act and/or move. In geared electromagnetic motors, which is the most widely adopted actuation solution for robots, there exists a fundamental trade-off between strength (at a high gear ratio) and interaction and back-driving capability (at a low gear ratio or in direct drive), which enforces a compromise in the performance of robot hands. Aiming to address this challenge, this project will fundamentally advance the actuation and control of robot hands and create and demonstrate a new type of robot hand that intends to have excellent manipulation dexterity and contact safety without sacrificing strength and hardware simplicity. This research holds a strong potential to improve productivity in warehouses and factories and thus mitigate current labor shortages, thereby benefiting society at large. In addition, this award will support the training of graduate and undergraduate students in research, incorporating the research into curricula, and outreach to K-12 students.The objective of this research is to create a new actuation, control, and learning framework for robotic hands to simultaneously allow for enhanced manipulation dexterity, contact safety, high strength, and hardware simplicity. The key idea is to create new robot hands driven by direct drive Vernier permanent magnet (VPM) motors (combining magnetic gears and motors) that affords high torque and excellent transparency simultaneously, which will expand robot hand operating conditions. Building on the hardware advances, this project will develop new sensor-less proprioceptive tactile estimation algorithms and new learning-based dexterous manipulation algorithms to fully exploit the new capability afforded by actuation innovation. This project will (i) create a systematic design framework for high-torque direct-drive VPM actuators optimized for robot hands; (ii) develop high-fidelity proprioceptive tactile perception algorithms to estimate contact location and forces; (iii) design and build a new robotic hand prototype and create learning algorithms for dexterous and safe manipulation. The performance of novel robotic hardware and integrated methods will be evaluated through experiments and comparison to state-of-the-art robot hands.This project is supported by the cross-directorate Foundational Research in Robotics program, jointly managed and funded by the Directorates for Engineering (ENG) and Computer and Information Science and Engineering (CISE).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.