This is a simulation project under the title "Swarm Exploration and Target Searching", carried out by 3rd year ECE students of Don Bosco Institute of Technology, Bangalore as a mini project.
Software used : Webots
Robot used : E-Puck
To locate a target in a multi-path (maze like) environment using a swarm of robots that co-ordinate and solve the task effectively.
In military situations were an Intel or an object of interest has to be located inside an enemy building or in a region affected by radiation, military personnel cannot be deployed. Instead a swarm of robots can be used as they are expendable and helps in locating the object faster when compared to a single robot.
- The arena is a square arena and it can be of any size.
- There are no obstacles in the path of the robot.
- Minimum distance between the walls is 1m.
- The distance sensors mounted on the robot can only measure upto 0.5m.
- The way the walls are placed or the expected distance/heuristic distance toward the target... are not given.
- The program is compatible with any size of the arena.
- More robots can be added by making some changes.
- Robots do not enter the already explored path.
- Priority is given based on the robot number to prevent conflict for same positions by multiple robot.
- Map data collected can be exported to a text file.
- Based on the exported data or directly from the program, the arena map can be recreated, with the starting positions of the robots and the target location.
- Continuous movement of the robot couldn't be achieved.
- The communication is centralized.
- The project is implemented on an existing robot (E-puck) (A better way would be to use a custom built robot).
- Target detection through camera of the robot is done through in-built 'recognition node' feature of Webots. (A better way would be to use OpenCV)