Agentic Twins: a solid foundation for AI-driven PLM
At the Future PLM Preview Forum, CTO Markus Herhoffer showed why powerful models alone are not enough and what AI agents need in order to act safely and transparently
by Christine Muggenthaler
What do AI agents have to achieve in complex industrial environments to be more than an answering machine, and instead act safely and transparently? That question was at the center of the keynote by our CTO Markus Herhoffer at the Future PLM Preview Forum on August 11, 2026.
His starting point: powerful models alone are not enough. Agents need a solid foundation of connected data, context, history and clearly governed capabilities for action. Only then do plausible suggestions turn into decisions that can be traced.
Agentic Twins: contextual representations of operational reality
In his presentation “Agentic Twins, a solid foundation for AI-driven PLM” Markus Herhoffer introduced the concept of the Agentic Twin: contextual, executable representations of operational reality that connect data across systems, time and organizational boundaries, enabling coordinated action.
For PLM this means a change of perspective: the digital twin is no longer just a representation to look at, but a working basis on which agents can assess relationships, estimate effects and initiate measures, verifiably and within defined permissions.
The central idea
Connect enterprise data.
Understand reality.
Act with confidence.
The event in review
Date: August 11, 2026, 15:00 UK, 16:00 CEST
Format: Future PLM Preview Forum, keynote 2
Talk: “Agentic Twins, a solid foundation for AI-driven PLM”
Speaker: Markus Herhoffer, CTO, EXP Software GmbH
Many thanks to the organizers and to all participants for the strong interest and the engaging discussion.