ENPH 353 Competition: An agent developed using Python to navigate through a simulated course, detect license plates of parked cars, and avoid pedestrians utilizing a state machine with image feed as inputs and velocity commands as outputs.
Figure 1: A bird's-eye view of the competition environment & our robot, Tofu.
The tasks our robot had to accomplish for the competition were:
- Drive withing the white lines of the road and navigate the parking lot.
- Avoiding collisions with pedestrians intermittently crossing at two cross-walks.
Figure 2: A pedestrian crossing from Tofu's point of view.
- Read license plates and parking stall numbers of parked vehicles.
Figure 3: A parked vehicle, with its parking stall number and license plate in view.
An overview summarizing our software architecture is included below:
