Development of a laser range finder based local-pose component for JAUS-compliant mobile robots in dynamic indoor environments

The ability to realize accurate and stable self-localization is one of the most important functions of an autonomous mobile robot. In this study, we estimate the local position of a mobile robot by employing a laser range finder using a new multiple-stage scan-matching algorithm. The proposed multiple-stage scan-matching algorithm utilizes indoor landmarks such as corridor walls and/or static objects to achieve accurate self-localization. On application, our proposed algorithm can recover mismatching errors that are caused by sensor noise and observations of moving objects such as moving humans. To confirm the validity of our algorithm, we implement a Joint Architecture for Unmanned Systems (JAUS)-compliant local-pose component. JAUS is a component-based interface for unmanned systems, developed by the United States Department of Defense. Employment of JAUS enables rapid prototyping and reduces the required development and implementation times. The proposed algorithm is implemented as a JAUS local-pose component documented in the SAE JAUS AS6009 Mobility Service Set. The validity of the proposed local-pose component was verified using an actual mobile robot in a dynamic indoor environment including people and unknown obstacles.