A scalable tree based path planning for a service robot

Path planning plays a vital role in a mobile robot navi‐ gation system. It essentially generates the shortest tra‐ versable path between two given points. There are many path planning algorithms that have been proposed by re‐ searchers all over the world; however, there is very little work focussing on path planning for a service environ‐ ment. The general assumption is that either the environ‐ ment is fully known or unknown. Both cases would not be suitable for a service environment. A fully known en‐ vironment will restrict further expansion in terms of the number of navigation points and an unknown environ‐ ment would give an inefficient path. Unlike other envi‐ ronments, service environments have certain factors to be considered, like user‐friendliness, repeatability, sca‐ lability, and portability, which are very essential for a service robot. In this paper, a simple, efficient, robust, and environment‐independent path planning algorithm for an indoor mobile service robot is presented. Initially, the robot is trained to navigate to all the possible desti‐ nations sequentially with a minimal user interface, which will ensure that the robot knows partial paths in the en‐ vironment. With the trained data, the path planning al‐ gorithm maps all the logical paths between all the des‐ tinations, which helps in autonomous navigation. The al‐ gorithm is implemented and tested using a 2D simulator Player/Stage. The proposed system is tested with two dif‐ ferent service environment layouts and proved to have features like scalability, trainability, accuracy, and repe‐ atability. The algorithm is compared with various classi‐ cal path planning algorithms and the results show that the proposed path planning algorithm is on par with the other algorithms in terms of accuracy and efficient path generation.