H ere’s a tale of two industries – cars and computers. In February, major car manufacturers such as Nissan and Honda began to warn shareholders that revenues were likely to fall significantly below expectations. And the reason wasn’t Covid-19 – well, not directly anyway: the pandemic had already significantly depressed sales in 2020. No, the problem was that manufacturers were now unable to make some cars because they couldn’t get the silicon chips (processors and other semiconductor components) needed to get the vehicles rolling off production lines. As a result, some factories were temporarily closing or being put on short time.
Meanwhile, in the same month, the computer industry was looking at a record year. Laptop sales were up 90% year on year. Tablet sales had recovered after a long slump. Even desktop computers and printers, for heaven’s sake, were flying off shelves and into delivery vans. So how did it happen that one industry struggled while another boomed?
The answer is that both had found themselves caught in a perfect storm that one had weathered and the other hadn’t. This storm bought three forces simultaneously to bear on an unprepared world: the fragility of a global supply chain on which both industries critically depended, the exigencies of US-China geopolitics and a pandemic that, more or less overnight, transformed the way large parts of the industrialised world worked.
Once upon a time, cars were made the Henry Ford way, revolutionary in its time, but involving holding huge stocks of components to feed a relentless mechanised production line. As Japan started to rebuild after the war, its leading carmaker, Toyota, came up with a more efficient way of making them. It came to be called the “lean machine” and a key feature of it was to hold very small inventories of components and instead have the necessary parts delivered just when they were needed for a particular assembly task. It was the beginning of just-in-time (JIT) manufacturing and it eventually became the way all cars were made because lower inventories meant lower manufacturing costs, better quality and higher profit margins.
But JIT critically relies on an efficient, reliable and robust supply chain. If the chain falters, then everything grinds to a halt. This applies whether the part is a gearbox or a silicon chip and over the last two decades chips, particularly in engine management units (EMUs), have become vital to the functioning of even the humblest petrol or diesel vehicle. We’re heading towards a future when cars will essentially be computers with wheels. But even now, if the relevant chips don’t arrive, then it’s crisis time.
Nissan may be a big cheese in the motoring business, but it’s a midget compared with Apple, Amazon, Google or Microsoft
The current distress of the car industry stems from the fact that the chips aren’t arriving – for several reasons. One is that there’s a global semiconductor shortage as a result of geopolitical rivalry between the US and China. This was triggered initially by the decision to exclude Huawei from western mobile networks. Another is that the computer and mobile industries, having seen what had happened to Huawei, began stockpiling chips on a massive scale.
A third factor is that when car sales began to dwindle in February 2020, the manufacturers reduced their semiconductor orders, leaving the relevant manufacturing capacity available to be snapped up by a computer industry struggling to meet an exponential demand caused by increased home working. And the coup de grace was that car manufacturers are relatively small beer compared with the electronics industry and so found themselves languishing at the back of a queue for a dwindling supply. Nissan may be a big cheese in the motoring business, but it’s a minnow compared with Apple, Samsung, Amazon, Google or Microsoft.
So we, not to mention the car manufacturers, have arrived at an interesting point. A huge industry built around the idea of propelling ourselves around via a series of controlled explosions, which, after all, is what an internal combustion engine is, needs to make a paradigm shift. VW, Ford, Mercedes, Volvo et al will need to become computer companies.
A few years ago, searching for a metaphor that would illustrate the change that’s coming, I came on two new cars side by side in a French carpark. One was a Porsche 911, a glorious, beautifully engineered triumph of baroque technology. The other was a Tesla Model S. And the metaphor that came to mind? On the left, in place of the Porsche, I saw a beautifully engineered Nokia phone, which was great for making calls and sending texts and not much else; on the right, the Tesla stood in for the first iPhone, which was basically a handheld networked Unix computer that could also, at a pinch, make calls. And we know how that story ends.
Nokia was a very interesting company that made great hardware. But one always had the impression that, at every critical moment in the development of one of its devices, the needs of the software, ie computing, invariably took second place. The hardware guys called the shots. Which is why the path that led the car industry to its current silicon deficit rang some sobering bells.
An interesting New York Times piece on how Nobel laureate Paul Romer became disenchanted with the tech industry. What took him so long?
Maker’s Schedule, Manager’s Schedule is a lovely essay by Paul Graham on why managers and makers inhabit different universes.