currently the actual speed information is obscured because of the averaging of speed across stops during OD reads
- ideally we'd be publishing/queuing whatever is the average speed before the stop and then publish also while it is stopped (somehow queue for later publishing instead of publishing right away!) and then resume data collection after the motor starts up again
-- so ideal scenario: see the dips of stops but don't just average across them
currently the actual speed information is obscured because of the averaging of speed across stops during OD reads
-- so ideal scenario: see the dips of stops but don't just average across them