Schneider Electric, NVIDIA and HPE Target Process Industries with Edge-Deployed Industrial AI Agents

The three companies are packaging HPE Private Cloud AI, NVIDIA AI models and Schneider Electric OT expertise into on-site agents for OT monitoring, cybersecurity and energy optimization.

Schneider Electric has announced a strategic collaboration with NVIDIA and HPE aimed at accelerating adoption of Physical AI in process and heavy industries. The first demonstration of the partnership is a set of Industrial AI Agents at the Edge, which the companies say will bring AI closer to the point where operational decisions are made.

The companies argue that today's Physical AI discussion is dominated by robotics, humanoids and autonomous machines, while process industries face a different problem. Their operators run complex, continuous operations in which reliability, safety, cybersecurity and energy efficiency are critical, and according to Schneider Electric no clear blueprint has existed for applying Physical AI in those environments.

The proposed answer combines HPE Private Cloud AI, co-engineered with NVIDIA within the HPE AI Factory with NVIDIA portfolio, with Schneider Electric's operational technology expertise. The intent is to let organizations deploy and scale AI applications at the industrial edge, keeping real-time intelligence near critical operations while retaining control over sensitive data, security and costs.

Three constraints targeted at the edge

Schneider Electric says the edge architecture addresses three issues that have limited industrial AI at scale:

  • Data sovereignty. Critical operational data is often too valuable, sensitive or regulated to move beyond operational boundaries. HPE Private Cloud AI is positioned as a secure foundation for keeping AI close to operations while maintaining control and compliance.
  • Cybersecurity. Added connectivity can introduce risk across industrial and critical infrastructure environments; keeping and processing data near its source is presented as a way to reduce the attack surface.
  • AI consumption costs. Cloud-based inference, data transfer and token-related costs can climb as usage grows. Edge deployment is said to reduce some of these costs and give greater control over where workloads run.

The first three agents

Customers deploy, manage and govern multiple Industrial AI Agents through Schneider Electric's EcoStruxure Industrial Insights apps, which the company says simplifies lifecycle management and security oversight. The initial agents cover three areas:

  • OT Monitoring Agents combine real-time operational data, equipment configurations, process topology, alarms, maintenance history and engineering context to assess the physical state of operations, identify emerging issues, reason about likely operational impact and recommend or initiate corrective actions, then verify that the intervention delivered the expected outcome.
  • Cybersecurity Agents correlate network activity, control system behavior, asset configurations, operational events and process context to detect abnormal conditions, assess potential impact on operations, recommend or execute responses and confirm that systems have returned to a secure and stable state.
  • Energy Optimization Agents integrate process data, energy consumption patterns, equipment performance, operating conditions and production requirements to identify optimization opportunities, recommend or implement adjustments and verify the resulting effect on energy efficiency, operational performance and sustainability outcomes.

An installed base at CMPC Chile

The foundation is already in production at CMPC Chile, part of Empresas CMPC S.A., which the companies describe as one of Latin America's largest pulp, paper and packaging producers. For three years the producer has used Schneider Electric's OTMS in production to monitor the health and availability of control systems across continuous operations, giving teams a unified view of control systems, servers, workstations and switches so specialists can spot issues earlier and set clearer maintenance priorities.

Matías Jory, digital and process manager in CMPC's industrial operations vice-presidency, said OTMS has improved visibility into control system health and availability, and that identifying issues early is critical to protecting production in continuous operations. He said the company sees potential for AI agents to build on that base, helping specialists investigate faster while people stay in control of critical decisions.

Gwenaelle Huet, executive vice president for industrial automation at Schneider Electric, said the next generation of industrial innovation will be driven by intelligence embedded directly in operations, and that combining automation, AI and edge computing expertise can help customers turn operational data into action. Sri Kodali of NVIDIA cited the combination of NVIDIA Nemotron open AI models and HPE Private Cloud AI with Schneider Electric's OTMS platform as the basis for agents that monitor, reason and optimize in real time. Alexia Clements of HPE said the goal is to bring AI closer to where data is generated and decisions are made while preserving the security, privacy and control customers require.

The companies describe the work as a foundation for progressively autonomous operations, framing it as automating the automation, with monitoring, reasoning and optimization capabilities expected to expand over time. The announcement did not include timelines, pricing or confirmed customer deployments of the new agents.

TB

Tomas Berglund

Tomas Berglund is an editorial analyst covering circuit breakers, switchgear, electrical panels, and protective devices in low-voltage distribution systems. His articles use rated voltage and current, interrupting capacity, short-circuit withstand, and the applicable scope of IEC 60947 to compare equipment without implying that any individual model is certified. He helps panel builders, protection engineers, and procurement teams frame coordination checks, request product evidence, and make safer specification and purchasing decisions.