Development of autonomous robot system for indoor messy environments

The purpose of this paper is to provide an example of system construction and experimental results for an autonomous robot applicable to indoor messy environments, commonly occurred in a disaster. Developed robot system is based on a testbed termed VSTR (variable single-tracked robot), which reveals driving mode transformation as occasion demands. Therefore, VSTR can realize obstacle avoidance as well as obstacle negotiation by itself if a proper autonomous system is added. Especially, in this study, an autonomous system for indoor messy environments is suggested with use of two sensors, PSD(position sensitive detector) sensor and LMS (laser measurement system), to measure the distance between a robot and surrounding obstacles. Under a proposed algorithm, VSTR chooses one of driving modes to cope with circumstances, and continues moving until the end of process. Experimental results for various target environments show that VSTR carries out autonomous navigation such as climbing stairs or coming out of area with disordered obstacles according to the proposed algorithm, and it is concluded that the results can be a promising alternative for civilian requirements of mobile robots in the future.