Every plant manager has lived this story: a pump starts vibrating oddly, and the answer to “what’s going on” is scattered across five different systems that don’t talk to each other. The historian has the trend line. A PDF buried in a shared drive has the maintenance history. The 3D model sits in an engineering tool nobody outside design ever opens. GIS has the location. Procurement has the spare parts list in a completely separate database.

Getting from “the pump looks off” to “here’s exactly what to do about it” has always meant a person piecing together disparate information whilst under time pressure.

That piecing-together problem is exactly what Cognite was built to solve, and on June 30, 2026, Schneider Electric agreed to buy the company for $3.1 billion in an all-cash deal. Once the transaction closes, Cognite will be folded into AVEVA and become a core part of the AVEVA CONNECT platform.

For operators across the Gulf’s industrial sectors, this is worth taking notice of. It’s not simply “another acquisition.” It’s a bet on where the entire industrial software market is headed.

Applications Used to Be the Prize. Now It’s the Data Underneath Them.

For years, industrial software vendors competed on the strength of their applications — whose SCADA was more reliable, whose historian scaled better, whose digital twin looked sharper in a demo. That competition hasn’t gone away, but a second, quieter battle has opened up underneath it: who owns and controls the data layer that all of those applications ultimately depend on.

Cognite’s answer to that question is Cognite Data Fusion — a platform that takes the siloed data challenge described above (process data, alarms, 3D models, maintenance records, procurement history, asset registries, GIS) and automatically contextualizes it into what the company calls an Industrial Knowledge Graph. Instead of five disconnected systems, you get one map of how a physical asset, its live data, its design history, and its maintenance record all relate to each other. That’s the foundation that makes it possible for both people and AI systems to actually trust the data enough to act on it.

By buying Cognite outright, Schneider Electric and AVEVA aren’t just adding a feature to CONNECT. They’re securing that foundational layer before someone else locks it down instead.

From “The Pump Might Fail” to “The Technician Is Already Scheduled”

There’s a useful way to think about how industrial AI has evolved, and where it’s going next.

First came predictive analytics: software that could tell you a pump was likely to fail, and roughly why. Then came prescriptive analytics: software that could also tell you what to do about it. The next stage — agentic AI — collapses the gap between insight and action entirely. An AI agent detects the anomaly, checks the parts inventory on its own, places the order, and books the technician, with a human reviewing the outcome rather than orchestrating every step.

Cognite’s Atlas AI was built specifically for this agentic mode of working, and it’s now part of what AVEVA brings to the table. Add in AVEVA’s earlier acquisition of TwinThread — AI models focused on optimizing entire processes, not just individual assets — and the direction is unmistakable: less dashboard-watching, more autonomous execution.

A Cumulative, Multi-Billion-Dollar Bet on Data Infrastructure

It’s worth putting this deal in context rather than reading it in isolation. Cognite brought in more than $170 million in revenue in 2025, with annual recurring revenue bookings growing 36% and fast adoption of Atlas AI. Layer that on top of AVEVA’s roughly $5 billion acquisition of OSIsoft’s PI System a few years back, plus its established Wonderware business, and the running total Schneider Electric/AVEVA has now put into industrial data infrastructure clears $8 billion.

That’s not a company dabbling in data. That’s a company betting its future on owning it.

Where Cognite Sits Once It’s Inside AVEVA

AVEVA already has CONNECT, its cloud platform for designing, building, and optimizing industrial assets. Once the deal closes, Cognite becomes the data and AI layer underneath it — the part most customers will never see directly, but will feel in everything they do see. AVEVA’s operational applications (InTouch, System Platform, PI), engineering tools (Unified Engineering, AIM), and digital twin interfaces stay exactly where they are; Cognite supplies the clean, unified data pipeline feeding them, and Atlas AI supplies the reasoning layer that lets them work together instead of in isolation.

What Changes for Customers on the Ground

Strip away the corporate strategy and here’s what it means for a plant or asset operator in practical terms:

• The integration burden shrinks. Today, connecting design tools to fragmented plant data usually means hiring consultants to build custom bridges between systems. A natively unified data fabric removes a lot of that work before it starts.
• Generative AI becomes usable on real infrastructure, not just in demos. Large Language Models (LLMs) are prone to making things up when they’re not grounded in structured, verified context. A knowledge graph that enforces that context is what makes it safe to point generative AI at an actual physical plant rather than a slide deck.
• Automation moves from “here’s a chart” to “it’s already handled.” Instead of a predictive maintenance dashboard flagging a problem for a human to act on, routine operational issues can be resolved by the system itself — freeing engineers to spend their time on the decisions that actually need a person.

Something to Weigh Before Your Next Platform Decision

If you’re evaluating central SCADA and Historian platforms right now, this acquisition reframes the questions you should ask. It’s no longer about “which vendor has the best screens today”, instead, it’s “which vendor is building the data foundation that will still matter in five years.” That’s the bet Schneider Electric and AVEVA have just made publicly — and it’s one worth factoring into your own roadmap.