The industrial world has a legacy problem. Factories and energy plants across the globe still depend on hardware-locked control systems built for a previous era: expensive to update, resistant to adaptation, and essentially incompatible with the artificial intelligence tools now reshaping every other corner of the economy. Nobody designed it this way. It simply calcified over decades of incremental investment, until replacing any one component meant risking the entire operation.
Schneider Electric and Microsoft are betting there is a better path, and they have picked one of the most technically demanding environments imaginable to prove it.
Green hydrogen sits at the centre of most serious industrial decarbonisation strategies, and for good reason. It offers a route to clean energy storage and feedstock for sectors that electrification cannot easily reach. The catch is cost. Producing hydrogen through electrolysis is electricity-intensive, and electricity typically accounts for more than 70 percent of total production expense. Shaving even modest percentages off that figure changes the economics of the entire enterprise.
Solid oxide electrolyzers represent the efficiency frontier of hydrogen production, but their operating conditions are punishing. Thermal balance, hydrogen flow, energy inputs and equipment health must be managed continuously and precisely. Human operators monitoring these systems in real time face a task that is simultaneously exhausting and error-prone.
Working with Indian green hydrogen company h2e POWER, Schneider Electric deployed an AI-powered control solution across a 20-kilowatt solid oxide electrolyzer system in Pune. The platform monitors and adjusts the electrolyzer continuously, managing every critical variable remotely and autonomously. Operators, freed from routine surveillance, redirect their attention toward higher-order decisions.
The system has now logged more than 6,000 hours of stable operation across partial and full-load conditions, placing it among the most durable autonomous electrolyzer demonstrations anywhere in the world. Stack wear dropped significantly. Electricity consumption fell by up to 10 percent. Translated to a commercial-scale 10-megawatt plant, that efficiency gain represents roughly 500,000 euros in annual savings, enough to materially shift the competitive position of green hydrogen against fossil-derived alternatives.
The h2e POWER deployment is a proof of concept, but the broader argument Schneider Electric is making runs deeper than any single project. At its centre is EcoStruxure Automation Expert, described as the first open, software-defined industrial automation platform, which decouples software from hardware so that automation applications can migrate across different equipment vendors and infrastructure generations without being rewritten from scratch.
Paired with Microsoft Azure’s cloud and edge infrastructure, the platform provides a migration path for industrial operators who cannot afford to halt production while modernising. The Industrial Copilot, an AI tool built into the system, automates the engineering tasks that typically consume the most time during modernisation: writing control logic, configuring systems and navigating technical documentation. Engineering teams report time savings of up to 50 percent, with changes that previously required weeks now completing in hours.
The ambition is to do for industrial automation what open software architectures did for enterprise computing: break the grip of proprietary systems, create genuine interoperability, and allow intelligence to accumulate rather than remain locked inside siloed hardware. The hydrogen plant in Pune is one data point. The pipeline of industrial operators sitting on ageing infrastructure represents the larger opportunity, and the larger problem waiting to be solved.
