There is no fee to register for the qualification stage, but you do need to register for the Tapia Conference in order to compete. Scholarships are available.
For qualification, you should use the Pioneers in Gazebo simulation using the Player interface.
The interfaces intended for use in the competition are position (movement/odometry), camera, fiducial extraction, and some sort of proximity sensing.
The bumper is the most coarse proximity sensor but we allow for more sophisticated laser range proximity sensing. Because this competition is geared for computer scientists, the general rule we have is that components must be commercial-off-the-shelf components that anyone could purchase.
For qualification, we chose the Player/Gazebo platform as the one to support. If you wanted to use MS Robotics Studio or another simulation platform, I think we can be convinced to let you use these alternatives, but the burden for a faithful and convincing implementation would be on you.
If you choose to use the Player/Gazebo system, we guarantee that the proper interfaces will be supported both in simulation and for Pioneer and iRobot Roomba/Create-based robot platforms. Chad’s group is currently working on documentation for assembly of a competition-quality robot using COTS components and the iRobot Create, which costs around $750 in parts SMURV.
Your score on the qualification task is based on the number of markers correctly identified. There are penalties for announcing a marker that does not exist.
You should submit through the submission page. Please send mail to robotics@richardtapia.org if you have not received the URL.
The conference will hold a poster session for robot competition participants to exhibit their work. Teams should prepare a poster illustrating their approach.
You can definitely use your own robot. We’ll provide Creates for those who want it.
* What are the consequences for not getting back to the starting position (no points; most of the points, etc.)?
Returning to the starting position is a bonus. The main task is to find as many markers as possible.
Each team is awarded points based on the number of markers located. In the case of a tie, time elapsed is used to determine a winner, so it is better to find more markers slowly than fewer markers fast.
The borders of the robot area will be published in advance but there may be other unknown obstacles.
Yes, the markers for robots will be distinct from the markers for the rescuees.
The robot will always start next to a wall facing away from a wall.
In the initial rounds, robots will run one at a time, so there will not be multiple home positions. If there are multiple robots, then each home location will have distinct coloring. I