Control and state estimation of a mobile assistive robot

Abstract

Assistive robotics technologies have the potential to help a wide range of people with motor and/or cognitive disabilities. In particular, mobile robotic manipulation systems can empower people with motor disabilities by aiding them in performing manipulation-based self care tasks. While mobile manipulation systems are physically capable of performing a wide range of tasks, their ability to carry out these tasks is limited through practical and ease of use constraints. This ease of use problem is only exacerbated in the context of assistive robotics due to the fact users of assistive robots can have physical or neurological disabilities. Further, home environments are almost always stochastic and dynamic which makes reliable mobility challenging. This thesis presents our approach to developing a mobile manipulation system designed for use by quadriplegics to perform manipulation tasks with a particular focus on the probabilistic algorithms and control theory used to drive both the mobile base and manipulators.