Improving the readability of decision trees using reduced complexity feature extraction
In this paper we present a teleoperation system that allows to control a group of mobile robots in a collaborative manner. In order to show the capabilities of the collaborative teleoperation system, we present a task where the operator collaborates with a robot team to explore a remote environment in a coordinated manner. The system implements human-robot interaction by means of natural language interfaces, allowing to teleoperate multiple mobile robots in an unknown unstructured environment. With the supervision of the operator, the robot team builds a map of the environment with a vision-based SLAM technique. The approach is well suited for search and rescue tasks and other applications where the operator may guide the exploration of the robots to certain areas in the map.