ONYX Insight has launched ONYX Cortex, a physics-informed AI platform intended to bring condition monitoring, performance analytics and maintenance workflows into one operational environment for wind asset owners and operators. The platform will debut at WindEnergy Hamburg 2026 and draws on more than 50,000 documented failure events, 140,000 turbine-years of operating history and data from over 180 platforms across 25 original equipment manufacturers. ONYX says Cortex can reduce false positive alarms by up to 95% and unplanned downtime by up to 70%. The launch addresses a practical issue for maturing wind fleets, where availability, maintenance decisions and commercial performance are increasingly interdependent.
One Platform For Asset Decisions
Cortex combines monitoring, intelligence and maintenance activity in a single platform, connecting turbine health data with the financial implications of operational decisions. Rather than leaving teams to interpret separate condition, performance and workflow systems, the platform is designed to turn an alarm into a decision that considers potential revenue, maintenance requirements and asset risk.
Its interoperability is central to that proposition. Cortex can bring together information from supervisory control and data acquisition systems and condition monitoring systems, including ONYX and third-party technology. Operators can access the intelligence through a web platform, mobile app, Model Context Protocol server or computerised maintenance management system, while API capabilities are intended to support integration with existing enterprise asset management and enterprise resource planning systems.
The platform extends ONYX Insight’s fleetMONITOR condition monitoring offering, which the company says is deployed across more than 32,000 turbines worldwide. Cortex takes that established condition monitoring expertise beyond individual component alerts towards a whole-turbine perspective, intended to shorten the route from identifying an issue to determining a response.
Physics Informed AI Foundation
The differentiating element in Cortex is its use of engineering physics alongside operational data. The platform is built around engineering models informed by more than 15 years of operational history, rather than relying solely on statistical pattern recognition. This approach is intended to give engineering teams explanations that can support confidence in the recommendations generated.
Cortex centralises monitoring across drivetrain, blades, blade roots, tower and foundation. That full-asset view is especially relevant for operators managing mixed fleets and adopting self-perform or hybrid operating models, where in-house engineering, operations and maintenance teams need a consistent basis for action.
Alexis Grenon, CEO at ONYX Insight, said:
“The economics of wind energy are increasingly determined by operational performance. Every hour of downtime, every unnecessary intervention and every missed opportunity to extend asset life has a measurable, often substantial, commercial impact. Turning that into revenue means aligning teams and data so operators have trustworthy intelligence they can act on with confidence. ONYX Cortex has been built to do exactly that: physics-informed AI, grounded in real-world data at an unprecedented scale, connecting condition, performance and workflow so every alarm becomes a revenue-optimised decision. Our objective is straightforward: give owners and operators clearer insight, greater confidence, and better outcomes across the entire operating lifecycle.”
Balancing Risk And Revenue
For wind operators, a developing equipment issue does not always demand the same response. A conservative decision to de-rate or take a turbine offline can protect an asset, but it can also remove output during periods when generation has greater market value. Cortex’s OPTIMISE capability is designed to weigh remaining useful life against revenue, operating expenditure, parts, people and equipment availability.
ONYX’s early modelling indicates that turbines with known issues could recover between 2% and 20% of their revenue potential by timing output for high-value market periods. The company says this can be achieved without increasing the risk of catastrophic failure or restricting an operator’s ability to respond to grid flexibility demand.
This positions operational data as more than a maintenance input. In a market shaped by ageing fleets, rising maintenance costs and pressure to maximise generation, the ability to make an informed judgement on risk, timing and commercial value can materially affect portfolio performance.
Operational Workflows In Focus
ONYX says its MANAGE capability has demonstrated improvements in operational traceability and field reporting efficiency, while reducing unplanned downtime by up to 70%. For one leading operator, the company reports savings of more than $38m over four years.
The company is already trusted by eight of the world’s ten largest wind operators outside China. Its technology is also selected as GE Vernova’s exclusive provider of onshore drivetrain condition monitoring systems, reflecting the importance placed on predictive intelligence as wind portfolios expand and mature.
By making condition, performance and maintenance data available through connected workflows, Cortex is intended to reduce the administrative and analytical friction that can delay action following an alarm. The platform’s focus on third-party systems is also notable, recognising that fleet operators often manage equipment and data environments from multiple providers.
A New Reliability Proposition
ONYX Cortex arrives with a clear emphasis on operational reliability as a commercial discipline. The platform’s combination of asset-wide monitoring, physics-informed AI and workflow integration aims to make turbine data more useful at the point decisions are made.
Its debut at WindEnergy Hamburg 2026 will give owners and operators their first opportunity to see how the platform applies its data foundation to reliability, maintenance planning and performance optimisation across wind portfolios. For a sector looking to extract more value from established assets as well as new capacity, the launch offers a focused proposition: fewer false alarms, better-targeted interventions and clearer links between technical insight and asset revenue.
