Would Apple Build A Car?

There are plenty of rumours flying around about Apple testing a car of some kind. Some claim it is a self-driving autonomous car. Others aren’t so sure. But is Apple building a car? Is this something they would consider?

Not Likely

While JLG’s analysis dismissing the idea (linked above) misses some important data (cars may not sell a lot of units, but they represent a healthy chunk of most modern economies in terms of dollars), it is also missing some important facts. Unlike electronic devices, cars are heavily regulated. You think passing FCC compliance takes some effort? Try passing crash testing with the NTSB. Oh, and then there’s the whole EPA compliance and testing. And making sure you are on-side with SAE standards. And the rules vary every so slightly between all 50 states in the U.S. and 10 provinces in Canada. And there are other federal regulations in the U.S. and Canada. And then you have the EU’s regional rules and the rules for each EU country. Rules upon rules upon rules, and some aren’t exactly harmonized.

The automobile is the single-most complex consumer product ever created. Don’t believe me? Consider that the average car has at least 2 complete computer systems, both running in real-time (one for the in-dash/in-cabin systems and one for the engine/transmission/brakes). A car is filled with many sensors (antilock brake sensors, radars for backup and blindspot monitoring, airbag sensors, seatbelt sensors, passenger sensors, temperature sensors, anti-knock, fuel-flow, fuel level, oil and coolant pressure, accelerometers for seatbelt pretensioners, etc). In addition to anywhere from 4 to a dozen radar transceivers, there is the AM/FM radio, GPS and Bluetooth radios, along with LTE and Wifi in some cars.

On top of all that, you have the safety systems: space frames, crumple zones, telescoping steering column, at least 4 airbags (and as many as 7 in some cases), seatbelts, head restraints, side-impact protection, safety glass. The engine is mounted such that it won’t intrude into the passenger cell, and instead help absorb impact.

All of this technology must work at a huge temperature range (from -50 to +50 Celsius). It must work in snow, hail, sleet, rain, sun, sand, dirt. It all has to keep working when moving the machine down the road at speeds up to 200 km/h (yes, Autobahn speeds). It has to fail gracefully so as to not kill the people inside the car, or around it. It has to sustain massive physical impacts, again keeping the occupants inside as safe as possible. Even your house isn’t this complicated. Your PC? Fuff. The PC is trivial compared to all of this.

Much More Work Than The Technology

This makes a car a daunting proposition. Getting into the business is not for the faint of heart. Consider Tesla Motors, Elon Musk’s venture. The company was founded in 2003, but their first car, the Tesla Roadster, didn’t hit the streets until 2008. It took 5 years of work, including over a year of testing and verification with the various regulatory bodies, before the car was ready for sale. It also took time to find a site for the factory, and then over a year just to get the tooling and processes in place. Sure, car factories are well-known entities, but they are also incredibly complex facilities. Setting one up isn’t trivial and it isn’t cheap.

Then, to top it all off, you have to deal with the laws in some regions (mainly Canada and the U.S.) to don’t allow manufacturers to sell directly to car buyers. That means setting up a dealer network, and the dealers cannot be owned by the manufacturer. You think Apple had to work hard to change how wireless carriers worked? Well, that is trivial compared to getting the rules for car dealers changed. Again, just ask Tesla how that is going.

Just to pile on, then you have to deal with recalls. Cars are one of the most closely monitored products when it comes to safety, and product recalls (and the cost associated with them) is part of the game. That, and the attendant lawsuits, add up to a significant cost in terms of dollars, as well as a big hit to your brand.

It doesn’t stop: there is the impact of a defect on your customer base. If iOS or OS X have a bug, at best people are inconvenienced. At worst, someone loses an hour or two of work. A defect in a car can cause injuries and death. And some of the things you think would be innocuous can turn deadly. Just look at the GM ignition switch issue: what looks like something easily avoided appears to have killed over 40 people and some estimates place that number at over 150. Not a big number in the larger scheme of things (between 30,000 and 33,000 people die in traffic accidents in the US each year). But a product defect that results in someone’s death puts the blame on the manufacturer, not the drivers. And companies pay for those recalls (GM is expecting a $1.2 billion charge related to just the ignition switch recall. The Takata airbag recall could cost several billion dollars across the entire industry).

I think you get the picture: building and maintaining a car company is a huge proposition. It isn’t for the faint of heart. So why would Apple bother? They probably wouldn’t. So what might they be doing?

Probably About Integration

What Apple is probably doing is testing new ways of integrating iOS devices into the car. This is more than just a link to Siri and the other functions on an iPhone. It is about integrating the phone into the workflow of operating a car. The best way to understand that is to actually test it while driving. To understand the environment, you have to analyze the environment. We all know, to some degree, what challenges a driver faces. But collecting objective data on the subject increases the understanding of it.

I also expect Apple may be testing out new sensors and other input that could be added to a car (or already exist) and trying to find a way to use that data in iOS in a useful way. As I mentioned, cars are stuffed with sensors, and the number of those in a car is only going to increase over time. Having a sensible way to use that sensor data, and present something useful to the driver is good. Again, the best way to understand it is to try it out.

Ultimately, this may be about Apple trying to integrate more completely into the dashboard, and possibly providing a new way of giving control and information to the driver. Garmin and Blackberry (with QNX) already have a significant presence in the car (and Garmin has a history in aviation and avionics, not just in-car GPS systems). Microsoft had a presence, but with Ford ending their relationship, they are now out of the car again.

Apple may simply be trying to get into the automotive game with more than just CarPlay. You don’t have to build a car to have a meaningful presence in the space, and being a provider (rather than a manufacturer) means Apple can get into the dashboards of many different cars with many different brands. That gives them greater reach than they can ever hope to get in building their own automobiles.