Currently, we don't have a clean way of terminating CV detections. This is bad if we have too many images and can't process all of them in time. To fix this, we'll kill the aggregator process on submit our cluster results. This will be done in two parts:

Pt 1: Modify aggregator to kill all current processes cleanly, then move from the current tick (CVLoiter) to the next tick (AirdropPrep).

Pt 2: Modify the route associated with it to call this aggregator end process.