This project is a team competition of navigating the Tribot on a table from a start position, determined by the initial placing of the Tribot, to a goal position, determined by a colored cross. The table will have obstacles that the Tribot should not touch. The objects that will make up the obstacles will be known but not their position. The obstacles will have an initial position but can also change during the race.
The setup will make an external camera available. In the competition, each team has 10 minutes to train their system on the camera. After this the team has 5 minutes to complete the course. Each team can restart their run any time. The fastest completion of the course without touching of any obstacle is counted for the final standing.
Each member of the team has to submit up to 4 page summary of their experience. This should include a brief description of the project and the approach of the team, but than each member can focus on the specific contribution. Please include comments of possible problems that were encountered and an evaluation of the performance at the competition. You should also add an outlook of possible directions for improvements. Finally, please comment where machine learning contributed or could contribute.
The project is basically the same as above with the additional challenge that the robot might be modified at the begining of the competition that might alter slightly the physique of little robi.
Please decribe your project individually in the format of a scientific paper as outlined in the quidelines. The page limit is 6 pages. This paper must include an overview of your apporach but can then focus on the specific contribution each member of the team made. I am especially interested in a critical analysis of your experiemce and a clear outline where machine learning did and could further contribute.
Submit to prof4155@cs.dal.ca with subject line Project by Monday Dec 10.