Human-machine collaborative paradigm for underwater defect detection and repair reinforcement in dam-reservoir systems

With the increasing service life of reservoir dams and the dynamic changes in their operating environments, the precise detection, repair, and reinforcement of structural defects in deep-water reservoir-dam systems have become critical challenges for ensuring the long-term safe operation of major water conservancy infrastructure. This study proposes a “Three-Integration and Six-Synergy” human–machine collaborative operational paradigm, which integrates multi-source data, diagnostic assessment, and operational control. The paradigm is supported by systematic restructuring across spatial hierarchy, workflow, and knowledge flow dimensions, enabling six synergies that coordinate underwater/shore-based, manned/unmanned, detection/repair, equipment/material/process, expert/field, and device/cable/obstacle operations. The development of a 300-meter-level modular multifunctional deep-water repair platform overcomes technical and equipment bottlenecks for fully unmanned, end-to-end operations in deep-water environments. Complementing this, an integrated digital twin control system based on human–machine collaboration enables real-time mapping and dynamic scheduling of operational equipment status, environmental parameters, and risk conditions. In a demonstration application at the Some Hydropower Station of Yunnan Province, the proposed methodology successfully accomplished a full-process underwater repair operation for simulated defects, validating the feasibility and engineering applicability of this technical approach. These findings provide a systematic solution for the deep-water repair of reservoir-dam systems, holding significant potential for advancing intelligent operation, maintenance, and safety assurance of major water infrastructure.