I imagine that would require much better cameras and cost more. On my car the tamper prevention measures make it hard to read from many angles as well.
I think the solution for the use case you describe is coming in 2020, where cars will be able to talk to each other and to road sensors. That technology will probably be used to collect mileage taxes and demand pricing as cars shift to electric, so identity will eventually be a component.
No need to wait until 2020. Your car is already being tacked by sensors that detect the unique ID's of the transponders in your tires (used to tell your car the tires' air pressure).
And since most people don't change their tires very often, with enough sensors deployed in various locations (parking lots, drive-throughs, gas stations, etc...) it is very easy to develop patterns of movement and build a profile of an individual person.
It gets even worse when you realize that the tire sensor data can be date/timestamp correlated with your phone's bluetooth radio. So a pattern can be developed for inside movement, as well.
Big retail chains are already doing this (the bluetooth part). No need to have the Radio Shack electric eye bing-bong on the front door to count customers anymore. Now you can just tally the bluetooth IDs and not only tell how many customers you're getting, but how many UNIQUE customers you're getting, which areas of the store they visit most frequently, how long they linger at various displays.
And yes, they know if you're just there to use the bathroom, thanks to the phone you take with you in there.
I wonder if some day this behavior profiling will allow stores to refuse you entry because you've been there 20 times in the last six months and never bought anything, but used the bathroom every time.
I worked for a company doing this in 2015-2016. Bluetooth isn’t the primary mechanism because your phone doesn’t passively transmit its Bluetooth Mac under normal circumstances. It’s primarily Wi-Fi that’s used for that (and no, the Wi-Fi anonymization work done by Apple and others wasn’t enough to block this). We did use beacons too for stores that had a store app (and thus could be setup to receive the beacon probes), but Wi-Fi was the primary triangulation method. We also fused data from stereoscopic vision cameras too to get a more accurate count (since some folks disable Wi-Fi or don’t have phones on them).
It’s a scary space, but that’s part of why I joined. The space needs more smart privacy focused engineers to make sure things are built respecting privacy (I rearchitected systems to where no one person could be tracked and only aggregate data existed).
I’d love to be utopian and say nothing like this should exist, but the reality is if we don’t build it, someone will, and likely do it poorly.
I imagine that would require much better cameras and cost more. On my car the tamper prevention measures make it hard to read from many angles as well.
I think the solution for the use case you describe is coming in 2020, where cars will be able to talk to each other and to road sensors. That technology will probably be used to collect mileage taxes and demand pricing as cars shift to electric, so identity will eventually be a component.