Apple, a company you may be aware of from their Pippin game console, until fairly recently had some very ambitious autonomous car plans. Ambitious in the sense that they were planning to build a whole car. Last year, those plans were drastically scaled back, and a recent New York Times article gives some interesting insights as to why.
We already knew that Project Titan, what Apple was calling their build-a-car project, was a total shitshow. In fact, in a previous story about Apple’s automotive ambitions, it was stated that:
Inside Apple, employees recently described the company’s efforts to build a car as a project lacking vision and in complete disarray. The recent layoffs, followed by Apple’s pursuit of talent and expertise from outside companies, are part of the company’s effort to “reboot” the project, said people with knowledge of the layoffs.
What this recent article gives now is a bit more insight into what Apple had been thinking during the development and eventual failure of Project Titan.
Five unnamed sources who worked on or were close to the project spoke with the Times, and what they described seems familiar to those who’ve been watching Apple for decades: novel, ground-up rethinking, a focus—almost fetish—for clean design, and a desire to radically transform the way people interact with machines.
This time, though, unlike for desktop computer interfaces, music players, or smartphones, it didn’t really work.
My favorite bit from the story is this part:
Apple even looked into reinventing the wheel. A team within Titan investigated the possibility of using spherical wheels — round like a globe — instead of the traditional, round ones, because spherical wheels could allow the car better lateral movement.
So, yeah, Apple was literally trying to re-invent the wheel. And not just any wheel, but Minority Report (or maybe iRobot?)-Audi-style spherical wheels.
Now, Apple’s not wrong, as such—spherical wheels absolutely would make for better lateral movement. But what Apple fundamentally wasn’t used to was building products that have ramifications beyond just being cool or working better or worse.
A car can kill you, and, generally, an iPhone or iPad really can’t. When it comes to taking a radical, out-of-the-box look at something like the wheels a car uses to go, steer, and stop, you can’t just focus-group test it like a touchscreen and decide if it’s a cool innovation or not.
You have to really, really test it, push it to its limits, see how it fails and when, and understand that if it does, people die. And that’s not even considering all of the DOT regulations and testing a spherical wheel system would have to go through before getting approval.
Other aspects of Project Titan seem to have had a similar approach:
From the beginning, the employees dedicated to Project Titan looked at a wide range of details. That included motorized doors that opened and closed silently. They also studied ways to redesign a car interior without a steering wheel or gas pedals, and they worked on adding virtual or augmented reality into interior displays.
The team also worked on a new light and ranging detection sensor, also known as lidar. Lidar sensors normally protrude from the top of a car like a spinning cone and are essential in driverless cars. Apple, as always focused on clean designs, wanted to do away with the awkward cone.
All of these things—motorized doors, control-free interiors with augmented reality displays, Lidar without the funny party hat—all of these things could be cool and novel and possibly even disruptive to what people understand a car to be, but Apple simply wasn’t in a position to execute any of them.
The fundamental reason why is that, as far as I can tell, Apple never really respected the very important differences between developing a piece of consumer electronics and a car. The difference isn’t one of scale, as I suspect Apple felt; it’s an entirely different thing.
A car isn’t like most consumer electronic devices like a phone or a laptop because it’s really a colony of many, many devices, all working together. A modern car has networks of computers, a mobile shelter complete with interior design and furniture, a complex prime mover that transforms chemical energy from a fuel tank or a battery into rotational motion, a system to suspend and guide the whole two-ton mass, and has to be able to be whisked around at high speeds, over bumps, through rain, and then left outside for years at a time and still work.
Apple has made some incredible products over the years, but a laptop is simply not a car.
I think Apple’s (well-earned) hubris made them gloss over this difference, and when they finally realized that they’re not a car company, it was really too late.
There were other disagreements within the team, like the question of how much autonomy makes sense:
There was disagreement about whether Apple should develop a fully autonomous vehicle or a semiautonomous car that could drive itself for stretches but allow the driver to retake control.
Steve Zadesky, an Apple executive who was initially in charge of Titan, wanted to pursue the semiautonomous option. But people within the industrial design team including Jonathan Ive, Apple’s chief designer, believed that a fully driverless car would allow the company to reimagine the automobile experience, according to the five people.
...but even this really seems like moot if they didn’t manage to accept the major, overriding challenge of mass-producing cars.
Now, they’ve accepted that they’ll let a company with decades of experience (or at least an understanding of the scales involved) building cars, and Apple will focus just on the underlying self-driving technology.
Apple is currently developing an autonomous shuttle based on a conventional car platform which they call PAIL (Palo Alto to Infinite Loop) that will serve as a testbed for their tech and shuttle their employees to work.
This is a more reasonable direction for Apple, though, as surprising as it is to say out loud (or at least type), I actually think Apple should look into releasing their own car.
Now, I don’t think the car should be totally developed at Apple—that’s proven to be a dead end—but I think they should leverage what they’re best at for a car of their own: design and user experience.
If Apple were to, say, buy chassis from Tesla and put their own bodies on the chassis and operating systems in that body, they could potentially have a winner.
It could be a car that looked and felt like an Apple product, but all of the important, needing-government-approval, hard-to-engineer mechanical parts would already be developed.
Really, Apple’s strength is in the front end, the part the consumer interacts with. When Apple switched from the PowerPC architecture to the Intel x86 architecture years and years ago, did anyone really care? Macs still felt like Macs, even if their internal guts were essentially the same as any Windows PC out there.
This is no different. Apple should make their car, but just the parts that Apple’s customers care about, which is not under the hood.
So, Apple’s not going to make their own car, which is a good idea. But, if they’re smart, they should at least look into making their own car on someone else’s car, because I think that could be an even better idea.