Living digital twins are transforming industrial plants into more dynamic, responsive systems, mirroring the adaptability of living organisms. These technologies are setting the stage for industries to operate in greater harmony, reflecting the organic rhythms of their facilities.
While many envision industrial settings as rigid and mechanical, they function more like living organisms. They have a pulse, a metabolism, and experience wear and tear over time. Subtle irregularities often signal deeper issues, much like symptoms in a living being.
From simulation to intervention
Digital twin technology has long aimed to replicate this complexity, but earlier versions lagged behind reality. They focused on simulation and observation rather than real-time intervention, often unable to proactively address detected issues.
Living digital twins, however, combine detailed digital models with the ability to act in the physical world. AI is central to this shift. Rob McGreevy, chief product officer at AVEVA, compares these systems to evolving organisms. They adapt to live conditions, providing visibility and control that reflect a facility’s organic rhythms, bridging previously isolated systems.
Agentic AI is now enabling systems to move beyond diagnostics, making autonomous adjustments to maintain industrial homeostasis. Yet human oversight remains key. Engineers define what “healthy” operation looks like and ensure AI systems stay aligned, monitoring their actions through living digital twins.
Living digital twins connect engineering data with real-time operational information through a digital thread, a connected record of an asset’s lifecycle. Advances in IoT sensors and software platforms have made this possible, streaming data into user-friendly interfaces and enabling predictive insights.
AI’s role and human oversight
Operators view digital twins through various lenses: sustainability, maintenance, or performance. The next step is allowing AI to act on insights from these perspectives, though manufacturers remain cautious about operational risks.
Manufacturers are eager for this progress. A 2026 Manufacturing Leadership Council survey found that 66% are using or planning to use agentic AI, with 40% expecting factory autonomy by 2030. However, Deloitte’s 2025 smart manufacturing survey highlights concerns, with 65% of executives worried about operational risks, including business disruption.
McGreevy notes that AI will likely remain in an advisory role for the foreseeable future, with full autonomy limited to repetitive tasks. Human operators are evolving, focusing more on defining operating limits and interpreting AI actions, ensuring alignment with desired outcomes.
Governance must also adapt. While regulators like the EU are setting rules, many organizations have yet to establish full safeguards. A study by AVEVA and IMD found that only 23% report significant changes in decision-making due to AI, with 43% noting little to no change. Only 27% reported substantial data sharing with partners, highlighting the need for improved collaboration.
Oversight varies across deployments. In some cases, operators approve every AI action; in others, AI operates within defined boundaries. McGreevy highlights examples of human-machine collaboration among AVEVA customers, demonstrating the evolving relationship between humans and AI.
The Evolution of Human Roles and Governance
As AI takes on more tasks, human operators are shifting their focus. Workers now manage multiple disciplines, spending less time on routine approvals and more on setting operating limits. Their expertise is vital for interpreting AI actions and catching errors early, ensuring systems remain aligned with operational goals.
Governance is also adapting, though slowly. Many organizations lack clear safeguards for AI decision-making. Regulators like the EU are stepping in, classifying AI safety components in critical infrastructure as high-risk and requiring human oversight.
Holistic Coordination and Future Resilience
Emerging technologies are enabling better coordination across industrial networks. Physical AI, such as systems that pair robotics and sensing technologies with AI-powered decision-making, is creating opportunities for more holistic coordination. “Manufacturing and critical infrastructure have operated as siloed systems for the last 50 years,” says McGreevy.
