DEVELOPMENT AND CONTROL OF A WHEELCHAIR-MOUNTED ASSISTIVE ROBOT WITH ADAPTIVE END-EFFECTORS FOR ACTIVITIES OF DAILY LIVING

Abstract

In the United States, approximately 6 million adults use a wheelchair for mobility, and a substantial proportion experience reduced upper-limb function that limits their independence. More than three-fifths of these wheelchair users need assistance in at least one activity of daily living (ADL). Although caregivers and family members provide invaluable support, there remains a pressing need for solutions that explicitly foster greater self-reliance. Robotic arms offer a promising avenue in this regard, with the potential to play a substantial role in promoting independence for wheelchair users. Within the field of assistive robotics, there is a notable scarcity of wheelchair-mounted robotic arms that combine sufficient payload, standardized ADL workspace coverage, and versatile grippers.Safety features such as power-loss braking and comprehensive collision monitoring are not universally implemented, and there is limited systematic data on ADL task performance across standardized workspaces. This dissertation investigates the design, development, and control of a wheelchair-mounted assistive robot intended to support ADLs. The research is structured around three primary objectives, each contributing to the overarching goal of enhancing independence and quality of life for individuals with mobility impairments. First, a user-friendly simulation and control platform is developed that integrates Webots with MATLAB to enable real-time joystick-driven testing of kinematics/dynamics, inverse kinematics, and self-collision checking. The framework includes a Bluetooth joystick interface, oriented bounding box-based collision monitoring, and inverse kinematics with singularity handling, and it is validated by showing close agreement between simulation and prototype on ADL workspace reach (91\% overall across multiple representative task workspaces around the wheelchair). These simulations also provide torque profiles used for motor selection. Second, guided by the simulation outcomes, a 6-DOF modular assistive robot has been designed and built. The system targets a 3 kg payload, achieves broad reach across the ADL workspaces, and incorporates foldability for stowage on the wheelchair. Lastly, the work advances a set of customized grippers and auxiliary end-effectors to address common ADL objects and actions --- including opening and closing doors and refrigerators --- within those ADL workspaces, with standardized quick attach--detach interfaces to support repositioning for left- and right-sided users.