Back in 2014, Apple launched a massive project, code-named Titan, to build a car. But a couple of years later, Apple is drastically scaling back its ambitions. According to Bloomberg, the company has given up on the auto-manufacturing dream entirely. Instead, it’s focusing on writing self-driving car software that could power cars manufactured by traditional automakers.
Silicon Valley moguls have gotten into the habit of jumping into new industries and quickly turning them upside down — as Apple’s iPhone did to the cellphone industry. But Apple’s car struggles are a reminder that not every industry is as ripe for disruption.
Manufacturing a car is really hard. And succeeding at it requires a different kind of corporate DNA than succeeding in the computing sector. Tesla has been slogging through these challenges head on for a decade, and profitability is still years away. Google seems to have decided that car manufacturing isn’t worth it — it’s focused on creating software that will eventually run on cars manufactured by others.
Now Apple seems to be seeing the wisdom of Google’s approach, implicitly conceding that car companies have unique capabilities that Silicon Valley can’t easily duplicate.
Why car manufacturing is so difficult
The best explanation I’ve seen of Silicon Valley’s struggles to build its own cars comes from industry analyst Edward Niedermeyer, whom I interviewed earlier this year.
Niedermeyer pointed out that cars are complex products with hundreds of moving parts, and customers expect them to work reliably for years, over tens of thousands of miles, and in all kinds of terrain and weather conditions. These challenges are magnified by the massive scale required to turn a profit in the car business — which means that every production mistake costs millions of dollars to fix.
That means that succeeding in the car business requires a degree of regimentation that’s rare in Silicon Valley, where innovation is often valued more than flawless execution. Car companies plan the manufacturing process in great detail and have extensive systems for detecting and fixing flaws. Building up the necessary knowledge, equipment, and organizational structure takes years — even decades — of expensive trial and error.
Niedermeyer offered an example of the kind of challenge Tesla has faced as it has tried to grow from a luxury carmaker into a mainstream one:
A great example is the problem of mold growing from inside the Model S’s roof, particularly in Norwegian cars. Because its large panoramic sunroof is difficult to manufacture and install to a precise specification, Model S roofs often leak. A lot of those leaks are so small that customers might not notice. But because Tesla used an organic-fiber pad at the edge of the sunroof, aggressive molds invade at alarming rates in certain climates.
Of course, an iPhone is also a complex product with many components that have to fit together perfectly. But cars are much bigger, have a lot of moving parts, include many different types of material, and are expected to work in punishing conditions ranging from blizzards to tropical rains. That means there are a lot more ways a car can break down, and there’s no substitute for years of experience testing cars in a wide variety of real-world conditions.
Tech companies and car companies are natural partners
Of course, just as technology companies can’t easily duplicate car companies’ manufacturing prowess, the same is true in reverse. Building great software requires a particular culture, structure, and set of skills that technology companies have and car companies mostly don’t. Which is why car companies and technology companies are increasingly rushing into each other’s arms.
And so far, it looks like technology companies have the upper hand. There are only a handful of major technology companies known to be working on self-driving car software — Apple, Google, and Tesla are most prominent — whereas there are lots of car companies looking for tech sector partners. So if Apple can develop self-driving software that rivals Google and Uber’s, it won’t be hard to find automakers eager to incorporate its software into their cars.
Focusing on self-driving software allows Apple to keep its options open. It can potentially work with several different car companies and focus resources on the partnerships that prove the most promising.
This leaves Tesla as the only Silicon Valley company trying to bring Apple’s traditional strategy of selling both hardware and software over to auto manufacturing. Tesla is betting that doing both will allow it to reimagine the way cars are designed and deliver a more elegant, compelling product. But there’s a big risk that CEO Elon Musk has simply bitten off more than he can chew.