Abstract: In response to issues such as coverage blind spots, low efficiency and high safety risks of manual inspections, lagging responses of monitoring systems, and information silos among subsystems in traditional fixed sensor gas monitoring network of coal mines, an integrated system deeply combining gas inspection robots with coal mine safety monitoring systems has been constructed. Adopting a technical approach based on multi-sensor data fusion, intelligent collaborative control, and real-time communication interaction, a hierarchical distributed fusion architecture has been designed. This architecture consists of a device access layer, a data fusion layer, a business logic layer, and an application presentation layer, achieving complementary advantages between mobile and fixed monitoring. In terms of data fusion, a multi-source heterogeneous data fusion algorithm based on credibility weighting has been proposed, enabling effective integration of data from fixed sensors, mobile robots, and environmental monitoring devices through spatiotemporal registration, feature extraction, and association decision-making, with data fusion accuracy reaching 99.8%. For collaborative control, a hierarchical control model has been established, a combined weighting method that integrates the analytic hierarchy process (AHP) and entropy weighting determines weight coefficients in multi-objective optimization functions. This approach achieves dynamic scheduling and path optimization for inspection tasks, reducing task response time by 40.2%. To address complex environmental factors such as underground dust interference and undulating terrain, an improved anti-interference simultaneous localization and mapping (SLAM) mapping algorithm and a robust multi-sensor fusion positioning strategy have been designed. By adopting a laser point cloud preprocessing method based on intensity filtering and adaptive Kalman filtering technology, positioning accuracy has been improved to 0.15 m, effectively mitigating the impact of environmental interference on navigation accuracy. The integrated system incorporates functional modules for real-time gas concentration monitoring, autonomous inspection of hazardous areas, rapid response to anomalies, and multi-level warning and linkage control, forming a comprehensive and responsive three-dimensional monitoring network. Field application tests demonstrate that the integrated system achieves a data collection success rate of 99.8%, reduces data transmission latency to 95 ms, improves monitoring accuracy by 15.0%, attains a warning accuracy rate of 95.8%, lowers the false alarm rate to 2.3%, achieves a system availability rate of 99.1%, reduces emergency response time by 55%, and enhances inspection efficiency by 60.0%.