Hacker News Reader

Comment by eschneider | Hacker News Reader

If you're going to try and track this stuff for real, keep in mind most devices like this use motion sensors to go into low-power mode when stationary and only transmit on the move.

View on HN

Also that some devices log data locally and require manual pickup + review to avoid detection. Also LEO have been known to temporarily disable such devices when people do scans to detect them for Undercovers.

>Also LEO have been known to temporarily disable such devices when people do scans to detect them for Undercovers.

Any more info on this?

Since I'm not seeing any other references, here's a timestamp for a YouTube video where an ex-undercover op is interviewed and such thing is mentioned: How FBI Undercover Agents Actually Work | Authorized Account | Insider https://youtu.be/h6au3ppTm7g?t=1123

> most devices like this use motion sensors to go into low-power mode when stationary and only transmit on the move

I've been working with (non-covert!) tracker devices for a project, and use exactly this approach, when stationary the tracker goes into low-power mode and sends position once every 12 hours to preserve battery life. When motion is detected, we send regular updates.

If low cost is the goal, consider a voltage measurement device. ICE engines have electrical systems that run ear 13V when the engine is on, and ~12.5-12.8V when the engine is off

That would require plugging into the wiring. At that point you no longer need a battery and can just use the car's power.

I'd be wary of draining the battery while the car is off. You don't want to prevent the car from starting

The ~10 or 20mA or so one of these things draws would take months to do that.

Untrue. Imagine a cute convertible car or sportbike in the snow country.

We talking MEMS/inertia detection, vibration detection or auto-geo-fencing?

The absolute cheapest thing is just to never update the position unless it significantly changed. Doesn't require anything except the GPS chip.

Bluetooth beacons would need to add an accelerometer, but that undermines their use in pinpointing an object at rest.

Probably the most effective technique for detection would be attained by spoofing the GPS signals, like the IRGC did to capture multiple US' drones?

https://www.gpsworld.com/gps-circle-spoofing-discovered-in-i...

I wonder how easily GPS can be spoofed, locally ...

https://rntfnd.org/2021/10/28/cheap-and-easy-gps-gnss-spoofi...

Seems someone already had the idea:

https://www.reddit.com/r/hardwarehacking/comments/10na5c8/sp...

I used to have a GPS repeater installed in our lab for RF testing. The FAA did not like it at all and threatened us with action.

Don't go spoofing or broadcasting your own GPS signals unless you have a decent legal team behind you.

Put it in an RF chamber and keep another GPS receiver outside the chamber some distance away to make sure it doesn't lose lock on the real satellites. That's your leakage canary.

We had one of those in an underground parking garage for autonomous vehicle testing at a previous job, but it was a naturally really well shielded room, and it was just repeating surface signals so no one would complain.

Using the GPS signal to detect motion is the most power-expensive path though.

The cheapest in terms of power consumption is a simple Accelerometer/Gyroscope component. The difference can be months or even years in longer battery runtime compared to GPS.

Whichever one is cheapest energy-wise. My guess is MEMS.

Sleeping the CPU until you get an interrupt from an IMU or simple motion detector is a common way to do this. It's not about being stealthy so much as extending battery life.

Reply on news.ycombinator.com