Unifying Rigid and Soft Grippers for Assistive Eating

Virginia Polytechnic Institute And State University, Blacksburg VA

Investigators

Abstract

Over one million American adults living with physical disabilities need help eating. Robots can assist these people by reaching for, picking up, and carrying bites of food to the user's mouth. Today's assistive robots pick up foods using traditional utensils (like forks), but these utensils fall short when the user wants to pick up small, slippery, and irregularly shaped foods. On the other hand, soft robotic grippers are strong candidates for grasping these foods --- but soft grippers may struggle to hold large, heavier items (such as a glass of water). This award aims to develop new fundamental understanding of object manipulation by combining soft, tunable adhesives and rigid, parallel mechanisms into a single rigid-soft robotic gripper. The team of researchers will explore how human operators use these grippers, and how these grippers can learn from humans to automate the process of reaching for and grasping items. The resulting rigid-soft gripper paradigm will expand the range of foods and other objects a robot can pick up. In addition to improving quality of life through assistive eating, this technology has applications in manufacturing factories, food processing, and fruit harvesting, where it can enable robotic arms to grasp objects of diverse textures, sizes, and shapes. To inspire and train K-12 students for future careers in engineering, the team will host live and remote robotic demonstrations where students control a robot arm and rigid-soft gripper. This project introduces a physics and algorithmic formalism for object manipulation that unifies rigid and soft robotic grippers along a continuous spectrum. The key insight is that --- instead of developing grippers that are either rigid or soft --- designers can leverage recent advances in active adhesives to coat rigid grippers with an array of soft materials. The team of investigators will: i) Characterize how best-case robots and everyday humans utilize rigid, soft, and rigid-soft grippers, ii) Learn from human inputs to autonomously and robustly pick up new items with rigid-soft grippers, and iii) Unify the spectrum of rigid and soft grippers under a physics equation that predicts gripping forces as a function of gripper design, object criteria, and human control patterns. The combination of a physics-based adhesion formalism with learned operator models will provide new insights into how adhesion phenomena and user inputs can enhance gripping capacity. The contributions will be evaluated through user studies where disabled and non-disabled participants teleoperate a robot arm and rigid-soft gripper. 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.