Neuro-Symbolic AI
Combining neural learning with symbolic knowledge representation and reasoning, with particular interest in Answer Set Programming and hybrid reasoning systems.
Personal Academic Lab
Research
My research explores how artificial intelligence can combine data-driven learning with explicit knowledge, reasoning and human understanding. The work spans theoretical AI, neuro-symbolic methods, human-centered AI and industrial applications.
Research themes
Possibility & Uncertainty
Possibility theory, possibilistic reasoning and approaches for representing uncertainty in neural and symbolic AI systems.
Knowledge Representation & Reasoning
Formal methods for representing knowledge, reasoning with incomplete information and connecting logical inference with machine learning.
Human-Centered AI & Rehabilitation
Human-centered and neuro-symbolic AI for rehabilitation contexts, including clinical documentation, participation and activity representation, environmental adaptations and practitioner-facing support.
Industrial AI, Digital Twins & Decision Support
AI-based decision support, anomaly detection and digital-twin prototypes for industrial systems, including energy production, wastewater treatment and inspection.
Current research
Possibilistic Neural Networks
Developing a formal connection between neural network outputs, energy-based representations and possibility/necessity measures, with logical constraints and reasoning.
Energy AI & Digital Twins
AI-based decision support and digital-twin approaches for monitoring, anomaly detection and risk management in energy and water-related industrial systems.
Tribo-AI
Hybrid computer-vision and symbolic reasoning approaches for analysing surface scratches and supporting interpretation of tribological measurements.
AI & Rehabilitation
Collaborative work with rehabilitation researchers and occupational-therapy practitioners on AI-supported clinical documentation, formal representation of activity and participation, and interpretation of environmental adaptations.
Collaborative work · Rehabilitation & AI
Selected collaborative work with Vera Kaelin and interdisciplinary rehabilitation partners explores AI-supported clinical documentation, ontology-based representation and neuro-symbolic reasoning. The work includes co-design of AI-supported rehabilitation documentation; image-to-text interpretation of environmental adaptations using CLIP, hierarchical classification and ontology-grounded generation; Transformer attention combined with Answer Set Programming for representing activity involvement; and a formal ontology-based treatment of activity and participation in the ICF.
Research material
This site contains selected research prototypes, experimental tools and project material. Publications and the complete academic record remain available through the university profile.
Research archive
Browse archive →Earlier projects and research experiments are kept as an archive because they document the development of ideas and methods over time. Some archived demonstrations may no longer be maintained.