AVEVA’s chief technologist on keeping humans in charge as AI starts adjusting industrial equipment

In an MIT Technology Review Insights podcast produced in partnership with AVEVA, the company’s chief technologist, Arti Garg, argues that industrial AI is entering a riskier phase. Foundation models, physical AI and agents can now automate more complex work in plants, power systems and mines, where an unexpected decision has physical consequences. Newer AI systems are harder to predict and explain, and their behavior can change as they are tuned, a different risk profile when the output is a set point on industrial equipment rather than text on a screen.

The piece comes from Technology Review’s custom content arm, not its newsroom. That does not make it wrong, but its success stories warrant outside confirmation.

Garg describes AVEVA pilots in which systems adjust equipment set points on their own instead of only suggesting them, with guardrails such as keeping automation inside a fixed operating band or limiting it to certain areas. She acknowledges the industry is still working out the human supervisor role and that guardrails for AI will not look like those given to people. That candor suits anything touching critical infrastructure, and banded, area-limited automation offers a practical template for letting AI act while humans stay accountable. Other industries are working through similar governance questions (Gilbert + Tobin).

Garg also describes AVEVA work with Idaho National Laboratory on AI for grid resilience as renewable generation grows, and says she chairs the IEEE P7100 working group developing a standard method to measure AI’s environmental impact across energy, resources, water and carbon.

Garg cites SCG Chemicals, a Thai petrochemical company, targeting around 99% plant reliability and an early-pilot return of almost nine times on its AVEVA platform. She also mentions an unnamed study suggesting industrial AI adoption rose nearly 78% over two years. These are vendor claims about its own tools, and buyers would benefit from independent data before relying on them.

Garg says close to half of the industrial workforce is set to retire within five years and frames AI as a way to capture that expertise. That is a real opportunity, similar to efforts to build organizational knowledge systems that learn from experts (organizational second brain). It also risks turning decades of judgment into a model that newer workers trust without the experience to question it, the failure she illustrates with a toy robot giving bad troubleshooting advice. Training staff to recognize when tools are wrong will matter as much as the tools. More in our AI coverage, enterprise coverage and robotics coverage.