A Dynamic Localized Adjustable Force Field Method for Real-Time Assistive Non-Holonomic Mobile Robotics

Abstract

Providing an assistive navigation system that augments rather than usurps user control of a powered wheelchair represents a significant technical challenge. This paper evaluates an assistive collision avoidance method for a powered wheelchair that allows the user to navigate safely whilst maintaining their overall governance of the platform motion. The paper shows that by shaping, switching and adjusting localized potential fields we are able to negotiate different obstacles by generating a more intuitively natural trajectory, one that does not deviate significantly from the operator in the loop desired-trajectory. It can also be seen that this method does not suffer from the local minima problem, or narrow corridor and proximity oscillation, which are common problems that occur when using potential fields. Furthermore this localized method enables the robotic platform to pass very close to obstacles, such as when negotiating a narrow passage or doorway.