The first 2 are readily addressable ... essentially we keep track of screens that are going off and on and car ignitions that are going off and on in some timeline then wait something like a week and go over it.
We shouldn't need to even correlate GPS ... I am presuming off/and time will be unique enough that obvious patterns will emerge. Then we can re-attribute accordingly.
This is spinoff of #94 - essentially we have:
The first 2 are readily addressable ... essentially we keep track of screens that are going off and on and car ignitions that are going off and on in some timeline then wait something like a week and go over it.
We shouldn't need to even correlate GPS ... I am presuming off/and time will be unique enough that obvious patterns will emerge. Then we can re-attribute accordingly.