Publication

Abstract : Natural disasters can wreak havoc on infrastructure and human life by creating dangerous scenarios where traditional rescue operations may be impossible. Autonomous mobile robots are critical to search and rescue missions in these scenarios because they can explore, map, locate, and navigate without a human onboard. An integrated approach utilizing sensor fusion for autonomous robotic perception and decision-making in unknown emergency situations is described in this research. In order to develop accurate environmental maps and navigate hazardous terrain, the system combines data from a LiDAR, IMU, web camera, and odometry. Webots is used to simulate a differential drive E-puck robot in a virtual disaster scenario. A range of path planning techniques are built and evaluated including A*, Dijkstra's, RRT, and the novel Adaptive Synergetic Path Optimization (ASPO). The sensor data downloaded in real time to MATLAB with the results indicating that ASPO allows for more flexibility during unpredicted events, and significantly improves navigation accuracy and stability along the path. This approach provides a flexible and robust solution for real world search and rescue scenarios, increases the operational efficiency of autonomous robots in dynamic disaster areas, and presents possible improvements for integration into mobile robots.