Urgent need for trustworthy Industrial AI- Solving the blackbox problem in critical environments
Dr. Gerhard Koch, Chief Software & Product Officer, Siemens Energy
Understand why Dr. Gerhard Koch, Chief Software & Product Officer at Siemens Energy, believes trustworthy industrial AI has become one of the most urgent challenges facing critical energy and engineering environments today.
The session explores the growing tension between AI’s ability to improve predictive performance, optimisation, and operational efficiency, and the risks created by black-box systems that lack transparency, explainability, and accountability. Siemens Energy highlights why hidden decision paths, hallucinations, and unverified outputs cannot be tolerated in environments where safety, asset reliability, and operational continuity are essential.
To address this challenge, Siemens Energy presents its “AI glass box” strategy, combining explainable AI, contextual AI, stability-tracking models, and deterministic safety guardrails to create systems that are interpretable, verifiable, and safe by design. The presentation also outlines the architectural layers needed to operationalise trusted AI at scale, from data governance and model trust layers to explainable intelligence and industrial safety controls.
The session ultimately positions trust not as a compliance exercise, but as the foundation for unlocking AI-driven industrial value, enabling organisations to scale predictive maintenance, autonomous optimisation, and operational intelligence with greater confidence and control.
