Summary

• Schneider Electric has agreed to acquire Cognite in a $3.1bn all-cash deal, with the business expected to sit inside Aveva.
• Cognite brings industrial data contextualisation, knowledge graph technology, and agentic AI tools for asset-heavy sectors.
• The deal shows industrial AI moving from analytics and pilots towards software embedded in energy, manufacturing, infrastructure, and plant operations.

Schneider Electric has agreed to acquire industrial AI software company Cognite in a $3.1bn all-cash deal, strengthening its position in the data systems that sit behind energy management, automation, and industrial operations.

The French group plans to buy 100% of Cognite Holding B.V., with the transaction subject to customary closing conditions and regulatory approvals. Once complete, Cognite will be integrated with Aveva, Schneider Electric’s industrial software business, and reported within its Industrial Automation division.

Cognite, founded in 2017, builds software for industrial data and AI. Its platform integrates, models, and contextualises operational, engineering, and enterprise data, giving industrial companies a clearer view of assets, production systems, and workflows. Schneider said Cognite generated more than $170m in revenue in 2025, with 36% growth in annual recurring revenue bookings, and employs more than 800 people across the Americas, Europe, the Middle East, and Asia-Pacific.

The deal gives Schneider a stronger position in one of the more commercially grounded areas of enterprise AI: systems that can work with operational data from plants, grids, refineries, factories, and infrastructure networks. Industrial AI has often been held back by fragmented data, old control systems, and the difficulty of translating models into decisions that can be trusted in high-risk environments. Cognite’s pitch is that AI needs a contextual data foundation before it can be useful in daily operations.

That differs sharply from the general-purpose AI tools being added to office workflows. Industrial environments tolerate less ambiguity. A model that misunderstands a document may create confusion; a model that misreads an asset condition, maintenance dependency, or plant history can affect output, safety, and energy performance. The commercial value of industrial AI therefore sits as much in data modelling, provenance, workflow integration, and governance as in the model itself.

Schneider already has a strong position in automation, energy management, and industrial software. Aveva gives it a platform across design, build, operation, and optimisation. Cognite extends that proposition into the data foundation needed for more automated decision support, including the use of knowledge graphs, generative AI, and agentic systems inside operational workflows.

Large industrial companies are also under pressure to reduce energy use, improve uptime, manage ageing assets, and adapt production networks to volatile demand and climate-related disruption. Many have invested in sensors, cloud platforms, and analytics tools over the past decade, but returns have often depended on whether data can be made consistent enough for use across operational teams. Schneider’s acquisition suggests the next phase of industrial software competition will be fought over who owns that contextual layer.

The deal also strengthens Europe’s industrial technology story at a time when much of the AI infrastructure debate is dominated by US cloud providers and chipmakers. Cognite is rooted in Norway’s industrial and energy software ecosystem, while Schneider is one of Europe’s most important automation and energy technology companies. Bringing the two together does not make Europe independent in AI, but it does give a European industrial group a deeper claim over the software stack used to optimise physical infrastructure.

Integration risk should not be dismissed. Industrial software acquisitions can look tidy on a strategy slide and prove awkward in deployment, especially where customers already use mixed estates of control systems, historians, enterprise platforms, and cloud services. Schneider will need to show that Cognite remains open enough to work across existing environments rather than becoming another proprietary layer in a crowded stack.

The acquisition is a bet that industrial AI will not be won by standalone models, dashboards, or pilots. It will be won by companies that can connect data from physical systems, understand operational context, and embed intelligence into workflows where production, maintenance, safety, and energy decisions are made.