Robots for Urban Search and Rescue (USAR) missions can perform numerous tasks that prove to be difficult for living agents (e.g. humans, dogs) without risking their lives. Some of those tasks require to fly over inaccessible areas, crawl into small spaces or run through hazardous environments with ease and communicate where, for instance, an injured person is found. As with every robotics application, realistic simulation test beds are advantageous for testing new ideas and algorithms in useful time with accurate results. Since, to the best of our knowledge, there are no publicly available ROS-based simulation scenarios for USAR that account with the chaotic spread of debris and objects following big catastrophes, we developed two of them which are now publicly available.
Each of the developed scenarios are available in three forms: easy, intermediate and hard. These “difficulty levels” were chosen according to the level of destruction, i.e., according to the difficulty of traversing the terrain on each level.
We invite everyone to contribute to this project, and have created two public repositories for it: one for the ROS-released code and another one that contains scripts to generate the scenarios.
Both repositories are available below:
- Github for ROS-released code: https://github.com/roboptics/destruction_scenarios
- Github for scripts to generate scenarios: https://github.com/roboptics/destruction_scenarios_generator
This scenario emulates a garage/parking lot that suffered from, for instance, an earthquake, which partially destroyed the columns, floor, ramps and stairs that compose it. The scenario also contains cars and people that can be the target of rescue missions.
This scenario represents an urban house/building that was also subject to an earthquake that partially destroyed the house’s walls and floors, along with its stairs. The scenario also contains people that can be the target of rescue missions.