Multi-modal control framework for a semi-autonomous wheelchair using modular sensor designs

This paper presents the hardware and software control framework for a semi-autonomous wheelchair. The hardware design incorporates modular and reconfigurable sensors and corresponding low-level software architecture. Two control schemes are discussed. Assisted control that augments the user inputs by providing functionalities such as obstacle avoidance and wall following. And, semi-autonomous navigation which takes higher level destination goals and executes a SLAM algorithm. We also propose an adaptive motion control with a online parameter estimation. The paper presents both experimental and simulation results.