Rate My Professor Andreas Pech

Prof. Dr. Andreas Pech is a professor in the Department of Computer Science and Engineering at Frankfurt University of Applied Sciences, affiliated with the institution since 1998. His field of expertise is Technical Computer Science. He teaches courses including Autonomous Intelligent Systems, Embedded Intelligent Systems, Robotics and Autonomous Systems, and Microcontroller Technology. Pech's research specializations encompass autonomous robotics, artificial intelligence applications, intelligent sensor technologies, and human-technology interaction. He is a member of the Research Center Future Aging, where he develops autonomous mobile assistance robots to enable elderly and needy individuals to live independently. This work involves autonomous navigation in variable and unstructured environments, independent grasping and handling of objects using robot arms, intelligent sensor technology and image processing, learning systems, machine decision-making, and human-robot interaction. Additionally, in the area of human-technology interaction, his research explores swarm intelligence in simple robots, emphasizing fault tolerance and robustness through decentralized algorithms for detecting system failures, synchronizing autonomous operations, and transferring tasks among robots with varying capabilities.

Pech contributes to the Competence Center for Applied Artificial Intelligence (ZAKI) at Frankfurt University of Applied Sciences. He has co-developed patented technologies, such as an ultrasonic sensor system for pedestrian detection in vehicles, granted Europe-wide in collaboration with Prof. Peter Nauth. His academic publications include 'On the Importance of the Newborn Stage When Learning Hierarchies' (2022, with D. Dobric, B. Ghita, T. Wennekers), 'Improved HTM Spatial Pooler with Homeostatic Plasticity Control' (2021), 'Scaling the HTM Spatial Pooler' (2020), 'The Parallel HTM Spatial Pooler with Actor Model' (2020, with D. Dobric et al.), and 'A maximum likelihood method for driver-specific critical-gap estimation' (2017). These works advance Hierarchical Temporal Memory (HTM) models, focusing on spatial pooler scalability, parallelism via actor models, homeostatic plasticity, and early learning stages. His research influences practical applications in robotics, AI for aging support, sensor-based safety systems, and Industry 4.0 initiatives.