Navigation of mobile robots in an intelligent environment with wireless visual sensors

Abstract

Today, most autonomous mobile robots rely on centralized perception and processing. They are, however, facing difficulties to be reliable and efficient due to uncertainties introduced from dynamic environments. In this paper, an intelligent environment with distributed visual sensors to support navigation and control of mobile robots, wireless mosaic eyes, is reported. The distributed visual sensors relieve the required massive intelligence to its environment. A multiple Bloom-filter scheme is developed for distributed storage of global routing information into wireless sensors. An error expectation based design for a multiple Bloom-filter is proposed, which exhibits superior performance to the conventional design. Considering dynamic navigation and control of a mobile robot guided by distributed vision, an active contour based scheme is developed to deal with path planning, trajectory generation, and optimal motion control. A working system has been developed to demonstrate the idea.