Assitive Feeding

Robot-assisted feeding promises to enhance the quality of life for individuals with mobility challenges by enabling independent meal consumption. However, current systems are limited to handling only curated, homogeneous plates, whereas real-world meals present a wide array of food items that demand varied, sequential manipulation strategies. This assistive feeding system addresses this challenge by combining the commonsense and few-shot reasoning of foundation models with a diverse library of parameterized skills to plan and execute efficient, user-preferred bite sequences. In evaluations spanning six realistic meal setups and feedback from 42 participants, this assistive feeding system demonstrated effective food pickup and seamless integration with existing bite transfer methods, even adapting across multiple institutions and robots. Its successful deployment in feeding a care recipient with severe mobility limitations further underscores its potential to transform assistive feeding in everyday settings.