An Integrated System for Perception-Driven Autonomy with Modular Robots

The theoretical ability of modular robots to reconfigure in response to complex tasks in a priori unknown environments has frequently been cited as an advantage, but has never been experimentally demonstrated. For the first time, we present a system that integrates perception, high-level mission planning, and modular robot hardware, allowing a modular robot to autonomously reconfigure in response to an a priori unknown environment in order to complete high-level tasks. Three hardware experiments validate the system, and demonstrate a modular robot autonomously exploring, reconfiguring, and manipulating objects to complete high-level tasks in unknown environments. We present system architecture, software and hardware in a general framework that enables modular robots to solve tasks in unknown environments using autonomous, reactive reconfiguration. The physical robot is composed of modules that support multiple robot configurations. An onboard 3D sensor provides information about the environment and informs exploration, reconfiguration decision making and feedback control. A centralized high-level mission planner uses information from the environment and the user-specified task description to autonomously compose low-level controllers to perform locomotion, reconfiguration, and other behaviors. A novel, centralized self-reconfiguration method is used to change robot configurations as needed.