Simplified EKF-SLAM by combining laser range sensor with retro reflective markers for use in kindergarten

Abstract

Recently in Japan, work contents imposed to childminders have increased due to several reasons such as variety of children's growth stage, requirement from parents, and administrative advices. To maintain good childcare service and to support kindergarten staffs, an objective monitoring system to watch over children is expected. In this paper, for the monitoring system, an automatic method to make an indoor map using the SLAM technique is discussed. To solve problem specific to the kindergarten use, using special landmark (LM) markers consisting of retro reflective material, an approach of a simplified EKF-SLAM which is effective by short-distance survey is proposed. Thorough simulation, the SLAM algorithm with an extraction method of the special LM from the laser range finder (LRF) sensor data was verified, and it was confirmed that the estimation error was as small as 0.57% against total length of robot locomotion. Same algorithm was implemented to an actual robot that was mounted with the LRF sensor, and practical effectiveness was validated. As a result, the ability of the presented SLAM with a margin of estimation error of 5.3% was also confirmed in the experiment.