Indoor Navigation With Human Assistance For Service Robots Using D*Lite

Service robots aim to help humans in their daily tasks at home or office. To accomplish some of these tasks, robots must be able to navigate intelligently from one location to another in dynamic indoor environments. The indoor navigation is a challenging task as many moving and movable objects can be found in the way, blocking the robot's path. D*Lite is a path planning algorithm that replans more efficiently in case the environment changes. However, depending on how blocked the robot is, replanned paths can be much longer than the original or, in the worst case, the target location can not be reached anymore. Therefore, we introduce an adaptation to D*Lite which allows robots to ask for human assistance to move objects on its surrounding when its path is blocked. The algorithm also computes whether and how long the robot can wait for help. Experiments were carried out using simulation and a real robot.