Autonomous Vehicles and the Semiconductor Industry: a Double-Edged Sword

A couple of weeks ago, I wrote about how automotive companies are going to need to change as the importance of electronics becomes the most important thing. But there are some potentially larger changes that will take place once fully autonomous vehicles are available, which is probably at least 10 years off. By autonomous in this context I don't just mean a car that can drive itself most of the time, I mean a car that can drive from one place to another without anyone in the vehicle.

Today, car sharing companies like Zipcar and CityCarShare allow some people to forgo owning their own vehicle. Typically these are people who live and work in cities and don't need a car to get to and from work. Uber and Lyft provide another way to get around, cheaper and more convenient than taxis. Of course Uber and Lyft don't just provide a vehicle, they provide a driver, too. But in an era of driverless cars, that would not be necessary. With a phone you could summon a car, have it take you to your destination, and then the car can go and seek another rider.

This potentially could mean that fewer people feel it is essential to own their own vehicle. Today, there are roughly 100M vehicles sold per year. This could drop, probably not precipitately but who knows. There is already widely acknowledged over-capacity in the automobile industry, especially in Europe. Most growth in sales today is in Asia, especially China. In the last seven years, growth was 16 million vehicles, but 15 million of them were in China. It is not expected to be quite so lopsided going forward, at least partially due to the softness of the Chinese economy.

Autonomous shared vehicles have the potential to disrupt the market. If large numbers of people who currently own a vehicle switch to just sharing one, then the market could shrink. The current overcapacity would become severe, and the fact that there are too many car companies and too many assembly plants would presumably lead to some sort of rationalization.

One counter-argument is that people still need cars to get to work. Even if they were shared, they would sit idle most of the day just like our own cars do today. It is one of the challenges of public transport. Trains somewhere like London go from the countryside into the city, then are warehoused in railyards until the evening when they go back. They are idle almost all of the time. If everyone was in an autonomous vehicle that would still largely be the case.

Some interesting symptoms of companies worrying about this come from General Motors. Last month GM invested $500M in Lyft. A couple of weeks later they also purchased the assets of Sidecar, the original smartphone ride-sharing service, which shut down. This could just be a patent play since Sidecar claims to own key patents that Uber and Lyft infringe but lacks the financial muscle to pursue, we shall see. Apparently GM will get a license to those patents but Sidecar, now a shell company, retains ownership.

When they made the investment in Lyft, the president of GM, Dan Ammann, said:

We believe that the first large-scale deployment of autonomous vehicles will be in this kind of on-demand, ride-share platform.

For the semiconductor industry, the enormous growth in automobile electronics could turn out to be a double-edged sword. On the one hand, selling a lot more silicon into each vehicle is a big opportunity. But on the other hand, if the number of vehicles sold drops due to the semiconductor content, then the opportunity may be smaller than forecast.

Thomas Kuhn, in his famous 1962 book, The Structure of Scientific Revolutions, invented the phrase "paradigm shift." It seems like autonomous shared vehicles could be a paradigm shift. But it turns out that scientific paradigm shifts like special relativity or plate tectonics don't really get accepted by the older generation being convinced by the new, but rather the older generation ages out and the new becomes the orthodoxy. As Max Planck (he of the ћ constant) supposedly said, "science advances one funeral at a time." (He actually said something much more verbose and it was in German, but I repeat myself.) In a similar way, it is young people who are most likely to adopt the changes first. Just as my children will never own a landline phone, their children may never own a car.