MARC - Modular Autonomous Adaptable Robot Concept

Abstract

The paper introduces a novel modular and adaptable payload concept for plugging in sensor and actuator platforms such as 3D LIDAR and visual sensor systems and robot manipulator and gripper systems. Integration, programming and operation of heterogeneous robot systems (such as mobile manipulators or robots in a machine network) are very complex tasks for plant operators. Heterogeneous system components have to be orchestrated (via proprietary interfaces) by higher-level control systems. Robot programs are created offline, in proprietary tools, and used through macros. System components are generally not compatible and interchangeable across manufacturers. These have to be programmed separately in manufacturer-specific tools. Cross-component debugging is difficult. The operation of complex systems is difficult, requires intensive training, and is currently limited to simple graphical user interfaces (GUI). As part of process optimization, it is usually necessary to optimize process points during the start-up phase. This requires highly skilled personnel capable of robot programming at the plant operators site. The combination of heterogeneous robot modules results in many new hazardous situations. It is therefore necessary to include an appropriate safety concept. The goal is to design a robot payload concept with a plug-and-play approach to be used as a modular and flexible unit. This shall reduce the effort for system integration and sensor calibrating significantly and provide a customized perception of the environment during certain work processes. The modular, autonomous and adaptable robot concept with several sensors and hardware components was implemented as a prototype on a rescue robot. The modules have already been integrated on other autonomous vehicles for exploration and dexterity tasks.