Artificial Intelligence is moving from experimentation to being an increasingly important part of industrial technology portfolios. As manufacturers look to apply AI across engineering, production, maintenance and operations, industrial technology vendors are responding by embedding AI into existing products, developing purpose-built applications, acquiring complementary capabilities and investing in the data and software infrastructure needed to support them. The result is not simply a wave of new AI features, but a broader evolution in how industrial technology is developed and delivered. We recently explored this shift in our report, The Role of AI in Transforming the Industrial Automation Technology Stack, examining how AI is reshaping industrial technology stack and the strategies vendors are adopting to capture the opportunity. This blog highlights some of the key trends on how vendors are advancing their products and portfolios as AI becomes increasingly embedded in industrial environments.
For many industrial technology vendors, the most immediate route into AI is to embed it within software that manufacturers already use. Rather than requiring customers to deploy separate AI applications, vendors are integrating generative AI, predictive analytics and more recently agentic capabilities into existing platforms covering product design, production planning, manufacturing execution and plant operations. This approach allows AI to become part of existing workflows while reducing the integration and adoption challenges associated with standalone applications. It also allows vendors to leverage their existing customer relationships, and industrial expertise while making AI accessible without requiring customers to adopt a completely new technology stack. For example, Honeywell Experion has incorporated agentic AI into its newly introduced Experion Cognition platform, which uses AI-powered agents for autonomous control room operations. AI is therefore increasingly becoming an intelligence and execution layer on top of existing industrial software rather than a replacement for it.
However, some vendors are also taking a slightly different route and developing AI applications specifically around individual industrial workflows. These include asset-performance copilots, reliability agents, engineering agents, alarm-management applications and autonomous operations tools. Siemens, for example, introduced the Eigen Engineering Agent (EEA) in 2026 as a purpose-built AI system for industrial automation engineering. Connected to Siemens' TIA Portal engineering environment, the agent is designed to plan and execute automation engineering tasks autonomously.
Building every component required for industrial AI internally can be slow. As a result, mergers and acquisitions are becoming another important mechanism through which industrial technology vendors are assembling the capabilities required for AI. The focus is increasingly extending beyond acquiring individual applications. Vendors are looking for connectivity, data contextualisation, industrial knowledge and workflow capabilities that can connect different parts of the industrial technology stack. Schneider Electric's agreement to acquire Cognite for USD3.1 billion illustrates this approach. Cognite provides an industrial data and AI platform designed to integrate and contextualise information from disparate industrial systems. The strategic value therefore extends beyond another software application to the technology can help connect industrial assets and operational systems with higher-level software and AI capabilities. The creation of Velotic provides another example of consolidation. In March 2026, TPG combined GE Vernova's former Proficy business with PTC's former Kepware and ThingWorx businesses to form a standalone industrial software company. The resulting portfolio brings together HMI/SCADA, MES, industrial data management, connectivity, IIoT and analytics capabilities.
These developments point towards a broader objective: creating technology portfolios through which industrial data can flow more easily from machines and control systems into operational applications and, ultimately, AI models and agents.
As vendors develop AI capabilities, they are simultaneously also investing in the data layer required to support them. Industrial data has traditionally been fragmented across OT systems such as sensors, PLCs, SCADA and historians, and IT systems such as ERP, PLM, CRM and maintenance applications. AI increases the value of connecting this information. An AI application may need access to machine telemetry, maintenance records, engineering documentation and production information simultaneously. Simply having access to individual datasets is therefore insufficient; the data also needs to be contextualised and linked.
This is driving investment in industrial data platforms, including IIoT platforms, data lakes and lakehouse, industrial data fabrics and semantic information models. Vendors including Siemens, Microsoft, AVEVA and Schneider Electric are building or expanding platforms that bring together industrial and enterprise data while adding AI capabilities. For example, in June 2026, Siemens expanded its Industrial Edge ecosystem through a partnership with HighByte. The HighByte Intelligence Hub can connect data from OT and IT systems, contextualise and transform it, and make the resulting datasets available for AI models, agents and applications. Siemens describes the combination as a unified data infrastructure for industrial operations.
For vendors, control of this data layer can create an important strategic position. The platform becomes the point through which AI applications access the information needed to understand industrial environments.
The AI transition is also reinforcing another long-running change in industrial technology: the move towards software-defined automation. Industrial technology vendors are shifting from selling standalone hardware products towards solutions in which software determines a growing proportion of functionality and customer value. Drives, robots, turbines and other physical systems can increasingly serve as platforms on which software enables configuration, simulation, optimisation, monitoring and AI-driven decision-making. This changes the relationship between vendors and industrial customers. Instead of the value of an industrial product being concentrated at the point of equipment sale, software and digital services can provide continuous functionality and updates throughout the asset lifecycle. Cloud computing, edge computing, digital twins and AI are accelerating this transition by allowing industrial assets to be monitored and optimised after deployment.
Another emerging strategy is the development of domain-specific industrial AI models. General-purpose AI models have demonstrated broad capabilities, but industrial applications often require more specialised knowledge. Engineering drawings, machine telemetry, industrial standards, maintenance histories and manufacturing processes contain terminology and relationships that may not be adequately represented by general-purpose models. Industrial AI therefore increasingly involves combining general AI capabilities with specialised industrial data and knowledge.
This could become an important source of differentiation for vendors. The competitive advantage may not come solely from access to an underlying foundation model, but from the industrial data, workflows, domain expertise and contextual information that surround it.
The industrial AI market is therefore developing beyond a contest between individual AI applications. Vendors are increasingly competing to establish the infrastructure through which AI can access industrial data, understand operational context and ultimately take action.
Embedding AI into existing platforms can accelerate adoption. Purpose-built agents can address specific workflows. M&A can provide missing pieces of the technology stack, while industrial data platforms provide the foundation on which AI applications depend. Software-defined architectures and domain-specific models can extend the role of AI further into the lifecycle of industrial assets and processes.
The longer-term implication is a shift in where value is created within industrial technology. As AI becomes increasingly accessible, the ability to connect data across the industrial stack, provide the right context and integrate intelligence into operational workflows may become as important as the AI model itself. Industrial AI is consequently not simply adding another layer of functionality to existing automation systems. It is encouraging vendors to rethink the architecture, ownership and integration of the industrial technology stack itself.
