UCLA

Recent digital technologies have envisioned a future ecosystem where objects are not merely passive entities but actively participate in and respond to the interaction between the user and the environment. While the majority of existing physical objects remain unsmart, researchers and practitioners have made efforts to incrementally upgrade legacy objects to perform digital tasks. However, there has been limited support for the automation of physical tasks, which is essential for users within certain demographics (e.g., the elderly and people with disabilities) or in scenarios of situational impairments.To bridge this gap, my thesis research aims to augment physical legacy objects with the ability to perform physical tasks automatically. My research explores two different types of augmentations: (1) involves augmenting the physical objects to perform tasks without external assistance, and (2) augmenting the objects to be more manipulable by robots. Specifically, I have introduced innovative methods for designing mechanical mechanisms that either empower objects to actuate on their own or enable actuation through an external robotic arm. A crucial aspect of enabling such robotic manipulation is a novel data collection technique I proposed, aimed at accurately estimating the pose of objects whose appearance changes during manipulation. Furthermore, given the diverse range of physical objects present in our daily lives, a central goal of my research is to democratize the design process, making it more accessible for augmenting these legacy items. In summary, my thesis aspires to enhance physical objects with the approach of robotic augmentations, smartening them to perform physical tasks for a more interactive and responsive future.