Abstract
This work presents a concept for autonomous mobile manipulation in industrial environments. Utilizing autonomy enables an unskilled human worker to easily configure a complex robotics system in a setup phase before carrying out fetch and carry operations in the execution phase. In order to perform the given tasks in real industrial production sites, we propose a robotic system consisting of a mobile platform, a torque-controlled manipulator, and an additional sensor head. Multiple sensors are attached which allow for perception of the environment and the objects to be manipulated. This is essential for coping with uncertainties in real-world application. In order to provide an easy-to-use and flexible system, we present a modular software concept which is handled and organized by a hierarchical flow control depending on the given task and environmental requirements. The presented concept for autonomous mobile manipulation is implemented exemplary for industrial manipulation tasks and proven by real-world application in a water pump production site. Furthermore, the concept has also been applied to other robotic systems and other domains for planetary exploration with a rover.
Highlights
The remarkable progresses in developing and interconnecting sensors, machines and people in the past years led to the idea of fusing high-tech networking know-how with stateof-the-art robot and automation technology into a smart manufacturing solution
The worker has to be supported by the system, solving the robotic part of the problems itself, which requires a superior level of autonomy
We present a concept toward fully autonomous mobile manipulation focusing on fetch and carry operations as important representative of industrial tasks
Summary
The remarkable progresses in developing and interconnecting sensors, machines and people in the past years led to the idea of fusing high-tech networking know-how with stateof-the-art robot and automation technology into a smart manufacturing solution. Having a mobile robot is crucial in order to work at different stations and tasks Coping with this problem requires sophisticated sensor data processing and perception capabilities as vital component of the system. We propose using this information during task training instead of utilizing complex reasoning approaches In this case, the worker has to be supported by the system, solving the robotic part of the problems itself, which requires a superior level of autonomy. Keypoints of our approach are: modularization: break down of the system’s functionality into small functional units; hierarchical flow control: coordination of modules for implementing complex behaviors; perception: sensing the environment to cope with uncertainties and changes; knowledge representation: abstract representation of the state of the task, the environment, and the system; and two phase approach: application of the given characteristics of tasks in the industrial domain. Conclusions and an outlook on future work are given
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
