Human-aware navigation for autonomous mobile robots for intra-factory logistics

Abstract

This paper presents a human-aware navigation system for mobile robots targeted to cooperative assembly in intra-factory logistics scenarios. To improve overall efficiency of the operator-robot ensemble, assembly stations and operators are modelled as cost functions in a layered cost map supporting the robot navigation system. At each new sensory update, the system uses each operator’s estimated location to affect the cost map accordingly. To promote predictability and comfort in the human operator, the cost map is affected according to the Proxemics theory, properly adapted to take into account the layout activity space of the station in which the operator is working. Knowledge regarding which task and station are being handled by the operator are assumed to be given to the robot by the factory’s computational infrastructure. To foster integration in existing robots, the system is implemented on top of the navigation system of the Robot Operating System (ROS).