Robot control architectures

In recent years, especially in the service robotics and as- sistance robotics domains, very complex systems have emerged. Due to improved sensor technology, faster and more energy-efficient control computers, new methods of perception, powerful methods of localization and map- ping, as well as cognitive system concepts, major progress could be achieved. Another aspect that supports the de- velopment of advanced robotic systems is the use of exten- sive open source robotic libraries. Frameworks such as the Robot Operating System (ROS) allow rapid realization of basic control systems. Furthermore, they have contributed to better reusability and integration of large collections of robotic methods in the community. Nevertheless, the increasing complexity of robotic systems is still a major challenge. Starting from a suitable embedded architecture via control software development supported by powerful frameworks to the control architecture, a suitable system design needs to be found. The goal is an efficient develop- ment process as well as many non-functional properties such as real-time guarantees, reliability, scalability, and security.