github.com/lidorshimoni/Fire-Fighting-Robot-using-ROS- ↗
Robot that explores a building while mapping and detecting fire hazards and survivors
Open this visualization on its own page →
Contributors
6
Lines of Code
590
From
2019-05-06
To
2020-02-26
About lidorshimoni/Fire-Fighting-Robot-using-ROS-
This project implements an autonomous fire-fighting robot that operates within the Robot Operating System (ROS) framework. The robot combines multiple robotics technologies to navigate buildings, map their layouts, and identify hazards. It uses SLAM (Simultaneous Localization and Mapping) to build a map of its environment while autonomously exploring unknown spaces, and employs computer vision with machine learning to detect both fire hazards and human survivors, marking their locations on the generated map.
The system runs in the Gazebo simulator and supports multiple control modes, including manual joystick control, autonomous navigation via move_base, and fully autonomous exploration. The implementation leverages several ROS components including frontier exploration for autonomous navigation, teb_local_planner and GlobalPlanner for path planning, and OpenCV with SSD (Single Shot MultiBox Detector) for object detection. The robot uses differential drive for locomotion and requires Python 3 with OpenCV among its core dependencies.