Welcome Prof. Dr. Jochen Müller to the THK-AI Research Cluster
The THK-AI Research Cluster welcomes Prof. Dr. Jochen Müller as a new member. With his addition, the cluster grows to 26 members and gains depth in building automation, digital twins, and semantic interoperability.
Short biography
Prof. Dr. Jochen Müller teaches and conducts research at the Faculty of Process Engineering, Energy and Mechanical Systems of TH Köln and is affiliated with the Institut für Technische Gebäudeausrüstung (TGA). There he heads the Cologne Lab for Building Automation and Optimization (COLABO), which is dedicated to the digitalization and automation of buildings and processes, from smart information modelling and modular plant automation to AI methods for process automation.
Research focus
At the centre of his research is the question of how heterogeneous building automation data can be made machine-understandable. Using AI methods from NLP and machine learning, building automation data are mapped onto a uniform information model, the Asset Administration Shell of Industry 4.0, creating a homogeneous semantic space. Large language models such as GPT-4o, DeepSeek-V3, Claude, and Llama are used for this purpose: the AI acts, in effect, as a translator between machines that previously spoke different languages.
In the research projects OptGA4.0 (BMBF funding line FHKooperativ, 2024 to 2027) and EcoTwin (EFRE/JTF programme NRW, 2024 to 2026), operating data are explored and modelled as digital twins according to this information model. OptGA4.0 optimizes engineering processes of municipal building automation based on standardized plant types and information models. The EcoTwin project links environmental data into a digital twin of urban green spaces. Building on this, applications for simplified building operation are being designed.
Connection to AI research
Prof. Müller’s work opens up an application field of immediate practical relevance for the cluster: the energy-efficient operation of buildings. Semantic interoperability, digital twins, and LLM-based assistance systems connect the cluster’s methodological strengths in machine learning, optimization, and NLP with technical building services, opening up prospects for joint projects at the interface of AI and building technology. We look forward to working together with Prof. Müller!