How to control a physical system, e.g., a robot arm, over a wireless link is still a considerable challenge: Questions like dependability and short latencies must be addressed; wireless resources need to be split between different physical systems; an urgent control need must be weighed against resources necessary to communicate over a bad wireless channel.
The project "Network-Informed Control, Control-Informed Networking" (NICCI, https://cs.uni-paderborn.de/cn/research/research-projects/active-projects/nicci/) addresses these challenges. It is part of the German Science Foundation's Priority Program "Cyber-Physical Networking" (SPP 1914, https://www.spp1914.de/home/). It is a collaborative effort between the control groups of Paderborn University (Prof. Daniel Quevedo) and Magdeburg University (Prof. Rolf Findeisen) and the Computer Networks group of Paderborn University (Prof. Holger Karl). The current position is hosted at the Computer Networks group, part of the Institute of Computer Science of Paderborn University.
NICCI's core idea is to improve the interaction between a distributed control system and the wireless resource management system. In particular, we are looking into learning techniques to find compact representation of a wireless channel's capabilities over short time horizons and derive best-possible decisions which transmissions should be undertaken between controllers, sensors, and actuators. This full-time position is currently fully funded for at least two years (extension possible pending grant decisions) at the E13 pay scale of German universities.