Dr Gerhard Koch of Siemens Energy, in this thought leadership presentation at Future Digital Twin & AI USA 2026, explores one of the most pressing challenges facing industrial AI today: how organisations can trust and scale systems whose decision-making processes are often difficult to understand.
Drawing on Siemens Energy‘s experience deploying AI across complex operational environments, Dr Koch examines the so-called “black box” problem and explains why transparency is becoming a critical requirement for AI adoption in energy, manufacturing and critical infrastructure. While AI has demonstrated significant potential to improve productivity, optimise operations and support decision-making, he argues that organisations will struggle to scale these solutions unless they can understand how models arrive at their conclusions and identify when those models can no longer be trusted.
The presentation explores several emerging approaches aimed at addressing these challenges, including model stability monitoring, explainable AI and contextual or physics-informed AI systems. Dr Koch also discusses the growing role of self-supervised learning, control agents and trust frameworks that can help organisations monitor AI performance, detect anomalies and maintain appropriate levels of human oversight in high-consequence environments.
As AI becomes increasingly embedded within industrial workflows, the need for transparency extends beyond technology alone to governance, safety, operational risk and workforce confidence. Watch the full presentation to learn how Siemens Energy is approaching trustworthy AI and why solving the transparency challenge may be key to unlocking the next phase of industrial AI adoption.
Speakers
- Gerhard Koch , Chief Software & Product Officer, Siemens Energy
