March 2nd, 2015

A Question of Frequencies

A few days ago, the Telegraph published this:

http://www.telegraph.co.uk/technology/apple/watch/11439847/Apple-Watch-will-replace-your-car-keys-says-Tim-Cook.html

"The watch is designed to be able to replace car keys and the clumsy, large fobs that are now used by many vehicles, Cook told The Telegraph. This could be a major development and will reinforce the view that Apple is circling the automotive market."

If there is something I hate almost as much as I hate paper, it is probably metal keys. They are the devil's own.

Somebody named Roger Lanctot published a rebuttal of this idea:

https://www.linkedin.com/pulse/apple-wont-make-car-toyota-already-does-roger-c-lanctot

If the Telegraph article was breathlessly positive about Apple and Tim Cook, Lanctot is the opposite.

"According to multiple industry contacts, but with a special contribution from Voyomotive, here is a summary of the requirements for a smartwatch/carkey replacement:

Connecting to cars will be different than controlling a car. Apple will need an OEM’s cooperation to control this functionality and most likely the car will need to be using CarPlay.
Apple will need to add hardware to the car and the vehicle will have to be designed around this hardware. Lead times are a consideration to implementation and market penetration.
Requirements include: Bluetooth communication with the vehicle “off;” however, it has to be very low power. System must be able to wake up the vehicle, support encryption and security. All of these are doable but are considerations to implement the system.
In the case of multiple users, Apple will need to differentiate permissions between different levels of users and Apple has traditionally NOT supported this type of functionality. For example, having different types of permissions for people who can enter the car vs people who can drive the car. (Once you are on an IPad – you have total access to everything.)
To reach the aftermarket and/or non-CarPlay vehicles, the Apple Watch will require additional hardware be added to a non-supported car to authenticate access and engine start. One model requires a custom install as used by ZipCar and other car sharing companies or a modular, plug and play system like Voyo."

So I find this a little suspicious. Why can't your Apple Watch talk to your Apple iPhone 6 or 6 Plus, and why can't the 6 or 6 Plus have a transmitter and software that does exactly whatever the fuck is happening inside the fob? Why should the cars change at all? The Watch could just trigger something on the phone that does what the fob does?

I believe the major sticking point would be in the hardware: can the Watch and/or Phone (probably the phone -- it's bigger and there's more space to put shit. The Watch is way cramped already) transmit on 315 Mhz or whatever the standard frequency is for key fobs? If it can, then Tim Cook is right: it's just a software problem, and getting the auto makers to cough up some APIs or whatever (here's hoping they _have_ APIs, and given what cell phones were like pre-iOS/Android, I bet they don't -- I bet they are written in assembly or worse) is all that really needs to happen to do the correct encryption and handshaking. There shouldn't need to be all that change on the car side. That's just silly.

So. Anyone able to figure out whether the transmitter(s) in the 6 and 6 Plus have the ability to transmit on the ranges used by existing car fobs? I'm failing miserably, but I'm bad this far down in the hardware weeds. If there has been a teardown and a part list, this should be an answerable question. Either Cook is right (the hardware is there) or not (there's no way that Lanctot is correct in asserting cars have to be modified -- you can reproduce the existing fob logic and transmitter in any number of physical formats).

I honest to goddess cannot figure out why Lanctot brought up Bluetooth. That doesn't seem to be anywhere near the current frequencies used by car fobs.

ETA: Interestingly, the BMW iRemote app will let me do a few things to the car by sending a signal to a server which then signals the car (this only works if you get the right option package and then pay for the subscription). You can turn on preconditioning (warm up the car and/or battery), flash the headlights and honk the horn (I found that last one hilarious when I found the feature) -- all intended to make departing in your car a little easier to set up a few minutes in advance. There's also a timer for preconditioning so you can warm it up while it is plugged into power. You can also _lock_ the vehicle, but you cannot unlock it. So if you walk away and forgot to lock it, you don't have to go back; you can lock it from your seat in the taxi or airplane or office. And if someone gets their mitts on your phone, they cannot unlock the car (which wouldn't do them a ton of good anyway, since they'd need to have the fob in the car -- or something that did what the fob does -- in order to drive anyway).

This should not be viewed as an argument about what the Telegraph/Cook suggests vs. what Lanctot asserts. It's just an observation about related existing functionality, and how one automaker chose to hobble that functionality. The BMW strategy appears to be: let's require them to have the fob if they want to use the car at all -- but let's not make them be within fob range in order to (a) find the car (b) pre condition the car or (c) lock the car. That way, you never have to dig the fob out of your pocket. Which is kind of nice. I think if I'd realized this, I might have quit locking the car when I walked away, and just did it via the phone, because digging the fob out sometimes requires me to take my gloves off to get it out of my pocket or whatever, whereas the phone works fine with the gloves (special gloves). Which this winter sort of matters.

Today's Project: a mail folder

Some time in 2009, I started planning our first trip to WDW. I created a new mail folder for email associated with that trip, and then just reused it without deleting the old mails for subsequent trips. I noticed the other day that there was ... a lot of email in there, and then it occurred to me that I could use this to actually assemble exactly who went with us on which trip when -- it's sad how fast the details fall out of my increasingly porous brain. So I pulled up a doc on google docs and I started going through the email, transferring information and then writing other things that I remembered along the way. I'm up to 2013, and 5 pages in. Ouch. Good news! I deleted several dozen messages.

ETA: This thing has really turned into a monster. I've got links to albums over on flickr, links into LJ, the stuff from the mail folder. I still need to finish going through the mail folder and do 2014 in LJ, but it's already at 14 pages. Yeesh.

On some level, I'm thinking of just turning this into Family Vacations, because I've already shoved in a couple Cape Cod trips and Efteling, because of connections to the Disney trips. And honestly, if I stick in Storyland (there's some of that in already) and Hershey, that's nearly everything anyway. The result would essentially be a picture- and blog- linked timeline of all the family vacations we took starting in 2009. With a little effort, I could probably back track to T.'s birth, or maybe even back to the Netherlands trip R. and I took before we got married.

What cars could have list -- still missing integrated child restraints.

http://jalopnik.com/ten-technologies-that-arent-in-cars-yet-but-should-be-1688816969

Whenever I see a list like this, I check it for whether they include integrated child restraints. These do exist in some passenger vans (5 point harness that you pull a panel away to expose, or put the panel back to cover it up comfortably for a rider who does not need it, but will instead use the regular 3 point). And yet no one ever talks about putting them in ordinary cars.

Makes zero sense to me.