The platform reported data cleansing accuracy of 92%, a 25% increase in data processing speed through edge-cloud coordination, and detection accuracy above 90% for small and medium-sized pointer and dial defects. Its dual-spectrum diagnostics achieved high precision in identifying equipment anomalies, and autonomous robotic intervention enabled safe emergency operations in inaccessible environments.
Large-scale renewable energy infrastructure is difficult to operate and maintain. Facilities such as photovoltaic plants occupy vast areas in remote locations, which makes manual inspections inefficient, and equipment such as substations, transformers and switchgear in offshore wind farms must work reliably in harsh environments far from shore. Inspection and maintenance traditionally rely on scheduled manual checks supported by basic sensor data, with infrared imaging and visible-light cameras each capturing only part of the picture. Operational data is often stored separately from inspection imagery, and many facilities run inspection robots and automated devices without centralized oversight, leaving data fragmented and assets underused.
China Huaneng Clean Energy Research Institute, with Huaneng Jilin Power Generation and China Huaneng Group, developed an AI-based platform to manage monitoring, fault detection and unmanned equipment across renewable energy sites. It links diagnostics and data processing through a layered architecture that connects sensors, AI models and operational workflows, and gives operations teams a consolidated view of asset condition across geographically distributed sites.
The platform supports three connected use cases. The first is fault diagnosis for offshore substations, where automated inspections cover the primary content of weekly maintenance personnel inspections at offshore booster stations; images captured by robotic inspection go through a deep learning pipeline that flags anomalies and pinpoints equipment-level fault locations. The second integrates infrared and visible light imaging into a unified diagnostic layer that merges visual observations with operational data, and when abnormal heat patterns align with irregular load behaviour the system flags the issue as higher priority. The third is management of unmanned equipment, with centralized oversight across group, regional and site levels and data cleansing algorithms that remove noise and correct inconsistencies before data is integrated into the broader system.
Across all use cases, edge-cloud collaboration allows data to be processed locally where needed while supporting centralized analysis and control. The system also supports autonomous intervention: in emergency scenarios, robotic systems can perform actions such as operating switches or valves in areas that are unsafe for personnel.