Arti Garg, chief technologist at AVEVA, says industrial AI is entering a new phase driven by foundation models, physical AI, and agentic AI. Unlike digital-only AI, industrial AI interacts directly with physical systems, where unexpected decisions can impact safety and critical infrastructure. Responsible deployment is essential.
Organizations must rethink data, governance, and human oversight. Industrial systems generate telemetry, service logs, and engineering documents across disparate sources. New technologies can connect and correlate this data faster, giving operators real-time diagnostic support. AI-powered robots could gather information in hazardous environments without endangering workers.
Greater autonomy demands new governance frameworks. AVEVA emphasizes security, efficiency, and human safety. Garg argues AI should augment rather than replace people in critical decision loops, with guardrails defining where automation acts and where human supervisors remain responsible.
Sustainability is also key. AI can help manage complex power systems as renewable generation grows, while organizations need better ways to measure AI's environmental footprint across electricity, energy, resources, water, and carbon. Garg participates in an IEEE working group developing standards for this measurement.
The next phase will bring industrial AI further into the physical world — autonomous robots, drones, and AI-assisted coding for domain experts. Realizing this potential requires rethinking business processes, establishing safeguards, and empowering experienced workers. "Autonomous systems will change the way we work in plants, power systems, and mining sites," says Garg. "That will make operations more efficient, safer, and more productive."
Source: MIT Technology Review AI · Summarized by HeadlinesBriefing