AI anomaly detection gains traction offshore as Aker BP deploys Cognite–NVIDIA solution
(WO) - Cognite has integrated NVIDIA’s NV-Tesseract time-series AI models into its industrial data platform, with Aker BP deploying the combined solution to enhance anomaly detection and predictive maintenance across offshore operations.
The integration links Cognite’s Industrial Knowledge Graph—designed to contextualize complex operational data—with NVIDIA’s AI models, enabling earlier identification of equipment issues and more actionable insights for engineers. The system is intended to move operators beyond traditional alarm-based monitoring toward predictive workflows that reduce unplanned downtime.
Aker BP is using the technology to expand monitoring across a broader range of assets, including wells, heat exchangers and other critical equipment. While the company collects large volumes of time-series data, only a portion is typically analyzed in real time. The new system applies AI models to continuously evaluate these datasets, identifying anomalies and flagging potential failures earlier in the process.
In addition to detection, the platform incorporates AI-driven workflows that connect model outputs with engineering context, including maintenance history and system documentation. This allows operations teams to transition more quickly from identification to response, reducing the need for manual troubleshooting.
The deployment reflects a broader shift in upstream operations toward digital tools that combine data integration with machine learning to improve asset reliability and operational efficiency. Predictive maintenance applications, in particular, are gaining traction as operators look to minimize non-productive time and extend equipment life.
By embedding AI models directly into operational systems, companies aim to improve decision-making at the asset level, especially in offshore environments where connectivity constraints and complex systems can delay traditional analysis.
The Cognite–NVIDIA integration will be demonstrated at CERAWeek by S&P Global in Houston, highlighting growing industry interest in scaling AI-driven monitoring and diagnostics from pilot programs to full-field deployment.
