Pairing Industrial AI Autonomy with Human Control

AI-powered 'living digital twins'—technology that recreates physical equipment exactly in a virtual space—have arrived on the industrial frontlines. As these systems enter a stage where they directly intervene in the physical world, they are emerging as a new technological trend in which autonomous operation and human control are realized together. According to content produced by MIT Technology Review Insights in partnership with software company AVEVA, a factory of cooling pipes and precision machinery now operates like a 'living system' with a pulse and a metabolism, with signs of equipment aging caught early through subtle anomaly signals.

Digital twins of the past could not perform real-time analysis, reacting several beats behind reality. They remained limited to simulation, observation, and planning of virtual scenarios. At the center of this change is AI.

Rob McGreevy, Chief Product Officer at AVEVA, likened the new AI-enabled generation of digital twins to a growing, evolving organism. Living digital twins, he explained, combine high-fidelity digital representations with operational and control systems capable of intervening in the physical world, enabling closed-loop execution.

The more advanced form of this technology is agentic AI. Agentic AI enables industrial systems to act autonomously, going beyond diagnosis or recommendations. At the same time, living digital twins provide human engineers with a means to monitor how AI operates and confirm that execution remains on track.

This content is not editorial coverage by MIT Technology Review; it was produced by MIT Technology Review Insights, the custom content division, in partnership with AVEVA. It is part of an industrial intelligence series addressing the theme of combining industrial AI autonomy with human control.