> A car on a busy highway needs to know which sensors belong to itself.

A sensor you were paired to disappears. Now a new sensor is showing up, and it sticks with you for an entire hour.

Sounds pretty easy for a computer to figure that puzzle out.

There are countless scenarios where cars are operated in close proximity for over an hour, like rural highway traffic or metro corridor traffic.

Every time a TPMS battery dies in these circumstances, the vehicle shouldn't pair with random TPMS sensors around it. Especially when we're talking about logic of a regulated safety system. It's a little better that it is deterministic, and follows an explicit pairing process.

> rural highway traffic

How big is the range on one of these?

> Every time a TPMS battery dies in these circumstances, the vehicle shouldn't pair with random TPMS sensors around it.

Random sensors around it that aren't already paired to their own car.

Also it could wait for you to complete an entire trip or two.

> Especially when we're talking about logic of a regulated safety system.

"Safety" in the sense that the little warning light usually gets you to do something about it eventually? Is this data going into anything where the correctness is a big deal?

> How big is the range on one of these?

As always with RF propagation, it depends. They're frequently in the 315Mhz band, so should be roughly similar to garage door openers, remote controls, etc.

> Random sensors around it that aren't already paired to their own car.

There's no handshake -- TPMS sensors are generally unencrypted broadcast devices. A car will see a lot of sensors. (and you can set up an antenna and track cars driving down your street) The "pairing" is simply the vehicle remembering which ones is theirs.

> Also it could wait for you to complete an entire trip or two.

It could. Now add the complication of: spare tires. And also, some but not all vehicles store the positionality of the sensor, so they can tell you which tire is low.

But if you're going to give the system so much hysteresis, you might as well just save the money and use the ABS-sensor based system that other vehicles use. These don't require any additional sensors or programming, but they are slower to react and don't provide pressure readings. The reason automakers use direct sensor systems is to provide a more direct and immediate reading.

> "Safety" in the sense that the little warning light usually gets you to do something about it eventually? Is this data going into anything where the correctness is a big deal?

It is a big enough deal that the reason many cars have them is to comply with the legal requirement that they have them. Before the light (and better cars have textual warnings), you'd have to manually check your tire pressure to identify an underinflated tire, leading to many people driving on them for extended periods of time and experiencing rapid unscheduled failures.

TPMS essentially automates people checking their tires because the reality is most people do not do a walk around on their car before driving.

In theory, we could use a dipstick in our fuel tank but most of us prefer an automated gage.

Even if you do a walk around, under-inflated tires are typically not distinguishable from normally inflated tires. Especially on today's cars with shorter and stiffer sidewalls.

I had a rental Mercedes with a leak in a tire recently... a tire was at something like 15psi but looked visually the same as the other tires. I absolutely do a walk around on all of my rentals and take pictures, but I would have had no clue if it weren't for TPMS. I would have driven it until it failed.

It had already failed. You had run-flats on the car. They’re still visually apparent.

Tires can have low enough pressure to affect vehicle handling without being visually low, you simply cannot measure tire pressure visually. That's why even tire shop workers use a gauge instead of eyeballing it.