Google and Westinghouse lean on AI to speed US nuclear plant builds
Summary
Google and Westinghouse have teamed up to apply AI and digital-twin technology to nuclear reactor construction and scheduling. The system couples Google’s AI models with Westinghouse’s WNEXUS 3D digital twin to break projects into millions of tasks, predict bottlenecks, re-order work in seconds and estimate costs. In a demo the tool reduced the cost of an air-handling equipment room by roughly $1m (around 25 percent) after optimisation.
The platform is moving from proof of concept into production. Westinghouse says AI-driven scheduling could halve traditional build timelines and expects ten planned reactors to be operational in five to seven years — though construction isn’t due to start until 2030. The programme also benefits from a reported $80bn backing announced earlier this year. Meanwhile, Schneider Electric is launching its own AI platform, EcoStruxure Foresight Operation, to unify energy and building systems management.
Key Points
- Google and Westinghouse combine AI models with Westinghouse’s WNEXUS 3D digital twin to optimise reactor construction.
- The tool decomposes builds into millions of tasks and finds optimal daily schedules to minimise delays.
- Demonstration claimed a near $1m (≈25%) cost reduction on a single component after AI optimisation.
- Supervisors can click an “AI Optimise” button to re-order work around delays in seconds instead of days.
- Platform now moving from proof of concept to production use, with claims it can halve traditional construction timelines.
- Westinghouse plans ten new reactors; construction starts 2030, with operations targeted in five to seven years after that.
- Schneider Electric unveiled EcoStruxure Foresight Operation as a competing AI-driven facility management offering (early adopters Q3 2026).
Context and relevance
This matters because large-scale power buildouts are expensive, time-consuming and critical to the expanding AI and datacentre ecosystem. Faster, more reliable scheduling and realistic cost forecasts could reduce overruns and accelerate capacity for power-hungry industries.
The use of digital twins and predictive AI reflects a broader industry trend: applying advanced analytics to complex engineering projects to regain lost institutional know-how, manage supply-chain uncertainty, and squeeze efficiency from labour and materials. However, critics caution that many scheduling gains could be achieved with traditional optimisation techniques — the novelty here is coupling those techniques with large-scale digital-twin data and rapid re-planning driven by AI.
Why should I read this?
Short version: if you give a monkey’s about where the power for AI farms will come from, read on. This isn’t just another enterprise AI demo — it’s about speeding up huge infrastructure projects, cutting costs and (maybe) plugging the energy gap for datacentres. If you work in energy, datacentres, construction or infrastructure planning, this shows where the sector is betting its chips: digital twins + AI for faster builds. If you like shiny buttons that rework schedules in seconds, this is your cup of tea.