Overview
A mobility system is developed for autonomous patrol and fire response in diverse environments using multimodal SLAM. Optical camera, thermal camera, and 3D LiDAR are fused to achieve robust localization and mapping under various conditions, including day, night, and low-light environments. An image-based fire detection model is also integrated to detect and continuously track fire, enabling active response. This establishes a foundation for reliable patrol inspection and fire monitoring in complex and dynamic environments.
Key contribution
▸ Robust perception and localization via multimodal SLAM: Stable localization and map generation across day/night and diverse environments through sensor fusion
▸ Fire detection and tracking method: Real-time situational awareness and response enabled by image-based fire detection and continuous tracking
▸ Scalable mobility system design: Designed for deployment on various mobile platforms, enabling a versatile autonomous patrol and response system
Impact
▸ Autonomous patrol robot capability: Stable perception and navigation across diverse environments
▸ Enhanced fire response capability: Reduced response time and improved accuracy through real-time fire tracking
▸ Improved unmanned monitoring and operational efficiency: Increased efficiency by minimizing human intervention
▸ Enhanced safety and cost efficiency: Reduced risk in hazardous environments and long-term operational cost savings