Real-world proximity sensing

[kevac]

Which sensor or combination of sensors could most accurately determine proximity of objects in open space over time, and the phones are in hands, pockets or purses while walking, bicycling or driving?

How would you prioritize the accuracy of these sensors from 0-5 meters (and how accurate are they at what ranges)?
- Proximity
- Microphone/sonar/ultrasound
- Light
- Blue tooth (assuming others also had their bluetooth on as well)
- GPS (assuming data sharing with others)
- Accelerometer/gyroscope/barometer (assuming data sharing with others, post-GPS established reference point)
- IP address (assuming data sharing)

[JanTangring]

Which sensor or combination of sensors could most accurately determine proximity of objects in open space over time, and the phones are in hands, pockets or purses while walking, bicycling or driving?

How would you prioritize the accuracy of these sensors from 0-5 meters (and how accurate are they at what ranges)?
- Proximity
- Microphone/sonar/ultrasound
- Light
- Blue tooth (assuming others also had their bluetooth on as well)
- GPS (assuming data sharing with others)
- Accelerometer/gyroscope/barometer (assuming data sharing with others, post-GPS established reference point)
- IP address (assuming data sharing)

Hi!

I notice that you got no answers here. Maybe you got lucky somewhere else and can share that here?

You posted in April – is social distancing possibly the reason for your interest?

[Jens Noriʇzsɔɥ]

Dear Jan and kevac,

time has apparently told that it is bluetooth. phyphox is also using it for quite a long time to identify the nearest external bluetooth sensors.

Once UWB (ultra-wideband) gets more established (Apple has included its corresponding U1 chip in most of its recent devices), this would prospectively be the way to go…

[kalaschnik]

Is there any news on fetching raw data from UWB devices/tags?

[Jens Noriʇzsɔɥ]

An intern tried to program a proof-of-concept in autumn. We have got the impression that the framework does not appear to be that ready: the code worked in theory (simulation), however, in practice we never obtained a stable connection. This has been with an iPhone and an Apple Watch, though, and I wanted to continue experiments with two iPhones between the holidays. Possibly, i(Pad)OS 15.1 or 15.2 include some bug fixes in this regard.

Unfortunately, AirTags cannot be used aside from Apple's “Find My”. I have not yet done any measurements on how precise its values are. My first impression has been that the direction is a bit off. This is also something to play with while having some days off.

I could not get a hands on any other UWB devices and the hardware developer kits appeared to be … rare and expensive. So, all in all it is pretty much early stage work in progress.

[Jens Noriʇzsɔɥ]

… and something that is really unpleasant: iPads are not built with a U1 chip… :/