Combine localization with estimating the position of a moving object, as well as decision making.
Have QR code landmarks serve as "bases". Have a large object or ball with QR code serve as "game ball". Robot will detect the distance away of the "game ball". Program (or train) bot that if "ball" is greater than certain distance away it will try to "run" to the next base.
Expand to always watch ball and retreat if it becomes closer. If we had time (we won't) we could turn this into a Reinforcement Learning task where the bot would learn when to run and not based on rewards of getting an out (i.e. if the ball is every within X distance and it isn't on a base, negative reward. Positive reward for each run scored)
Combine localization with estimating the position of a moving object, as well as decision making.
Have QR code landmarks serve as "bases". Have a large object or ball with QR code serve as "game ball". Robot will detect the distance away of the "game ball". Program (or train) bot that if "ball" is greater than certain distance away it will try to "run" to the next base.
Expand to always watch ball and retreat if it becomes closer. If we had time (we won't) we could turn this into a Reinforcement Learning task where the bot would learn when to run and not based on rewards of getting an out (i.e. if the ball is every within X distance and it isn't on a base, negative reward. Positive reward for each run scored)