Goals re-visited:

1) Detect ball and goalie using computer vision

• We were able to successfully detect the ball and goalie using our camera. Initially, the centroids for the ball and goalie were shaky, but by the end of our fine-tuning we produced stable centroids, accurately demonstrating the location of the ball and goalie

2) Actuate on goalie using motor

• Our goalie was able to move at our desired speed (>10 cm/s) using our motor

3) Plan where the ball will be in the future based off of past ball positions

• Our planning algorithm could successfully predict where the ball would be in the future based on the past 10 recorded positions of the ball. We tested this by slowly moving the ball along a straight line and saw the goalie move to the endpoint of the line we were tracing before we reached that endpoint with the ball.

4) Block >75% of shots

• During our testing, we took 20 on-goal shots (and many, many off-goal shots) against our goalie. Pickford (as the goalie is lovingly known) was able to block 18 of these shots, meaning we met this goal at a 90% success rate for blocking shots.