Towards a low-cost framework for Intelligent Robots

Abstract

Intelligent Robots can take advantage of a distributed, web-based information system deployed in the cloud to perform high level tasks. This paper proposes a robotic control framework suited to be used by low-cost robots, performing teloperated and/or autonomous tasks. The first part is dedicated to the development of an Android based robot control framework. This framework connects to specific low-level controllers that were developed for multicopters, wheeled/tracked mobile robots. The second part is dedicated to "place" the Android based robot in the cloud. There, the robot can perform Cloud based highly automated cognitive tasks in order to optimize their use and take best advantage of previous knowledge models, e.g., objects databases or 3D world models. Also, the robot can be controlled remotely using a classical teleoperation mode, using wifi networks. First experiments are presented when a tracked robot is performing surveillance tasks, while its state can be changed to teleoperation/videoconferencing mode, while interacting with a reasoning engine in the Cloud.