Human-Centric AI Partners for Operational Excellence

Challenge

Industrial production still relies heavily on expert operators whose decisions are guided by tacit knowledge, gut feeling and human senses. While this expertise is invaluable, it is also fragile: it is subjective, difficult to transfer, inconsistent across teams and often lost when experienced operators leave. As production systems become more complex and dynamic, novice operators struggle to reach expert performance, tuning decisions are hard to explain or justify, and companies face downtime, variability and missed optimization opportunities.

Project goal

Via this Strategic Basic Research project (ThoughtPartner_SBO), we aim to enable human-centric, explainable decision support for dynamic production systems. The project develops a “partner of thought”: an AI-driven system that captures and formalizes expert intuition using multimodal sensing (mimicking sight, sound, touch and smell), hybrid digital twins and probabilistic cognitive models.

By transparently explaining its suggestions and learning from past human–machine interactions, we help novice and expert operators make faster, more confident tuning decisions, align practices across teams and continuously optimize process performance. This results in reduced downtime, improved operational efficiency and scalable retention of expert knowledge.