I’m glad to see that there’s somebody out there who thinks the same way I do about autonomous vehicles. As it turns out it’s Jeremy Bowman at the Motley Fool:
While the excitement over self-driving cars is clearly palpable, there are a number of reasons the autonomous vehicle revolution may not happen as soon as some think.
There are 253 million cars on American roads today, nearly one for every person in this country. The average age of those vehicles is 11.4 years. While it’s easy enough to envision an urban millennial foregoing car ownership and depending on apps like Uber and Lyft, as many already do, making that system work for the majority of Americans will be much more difficult and slower to take hold.
The average value of a used car is now close to $20,000. Though the average age of a used car sold is just around five years, it helps show the value of the cars on the road today. Even if we assume the average value of a used car is $10,000, that would put the value of all of the cars in the U.S. today at $25.3 billion. After real estate, there is no personal property Americans have more money tied up in than cars, and they aren’t about to abandon that kind of money to take advantage of self-driving vehicles. The added cost of self-driving functionality may also dissuade potential buyers. The need to turn over all of that inventory means that once autonomous vehicles become widely available, it will still take about an additional 15 years for the fleet of cars on the road to turn over. So, even if self-driving cars are in mass production by 2020, the complete shift wouldn’t take place until 2035.
The high-profile fatal accident this summer involving a self-driving Tesla underscored a problem with the transition to autonomous vehicles. Since most auto accidents are caused by human error, autonomous vehicles have the potential to make driving much safer than it it is today, but the technology just isn’t there yet. Deploying it is likely to reveal similar faults to the one that caused that accident this summer, making the transition a tricky one for regulators and insurers.
The Obama administration endorsed autonomous vehicles and issued a set of guidelines surrounding the burgeoning industry, but toeing the line between ensuring public safety and allowing innovation may not be so easy. Though autonomous driving technology is expected to lower insurance rates, the question of who is at fault in an accident in such a vehicle remains an open one. Some have posited that automakers will offer their own insurance to cover the vehicle.
Plenty of questions still need to be answered in the regulatory environment, but an entrenched car culture and the bureaucracy surrounding the auto industry won’t be undone overnight. In the self-driving revolution, the technology may only be the first step since consumer acceptance and inventory replacement will take several years to overcome before autonomous vehicles become mainstream.
The reasons that the adoption of autonomous vehicles is likely to be slower than many think outlined above include reluctance on the part of consumers to destroy their own wealth and regulatory barriers, both of which I’ve mentioned in the past. There are four other reasons you might want to consider: the 90% rule, security, liability, and lack of value.
The 90% rule, well known in development circles, goes something like this. When a development is 90% working, you’ve completed 10% of the effort. All of those concept cars about which you’ve read glowing articles? That’s 90% of the effort. Completing the remaining 10% is likely to take much longer than anybody thinks and possibly much longer than is worth completing.
Security is another big issues. It’s one think when hackers break into your computer. It will be another thing entirely when they break into your car.
The insurance and regulatory issues mentioned above don’t touch on the issue of liability. Once upon a time there was a computer company (Burroughs) that was driven out of business by lawsuits due to bugs in their operating system.
My guess is that there are probably 10,000 lawyers out there ready to go to court to argue that any problems with an autonomous vehicle are failures of workmanship. Another way of saying that is that even Google with its notoriously deep pockets is about one class action lawsuit away from bankruptcy. It’s something they’ve never faced with the search engine and marketing tools that are their core business. They’re smart guys. They probably know it. They want to promote the technology they’ve developed, not put autonomous vehicles on the road.
The final issue is consumer benefit. We know how the technology is supposed to benefit manufacturers. What’s the benefit to consumers? Especially consumers who already own old-fashioned, antiquated non-autonomous vehicles.
Here’s a test case. Fully autonomous vehicles have been practical for decades in one area that is much, much simpler than automobiles: railroad trains. Why aren’t all choo-choos autonomous? I’ve listed the reasons above.
Consequently, I think that we’re going to see crash protection, unassisted parking, and other applications of smart vehicles that haven’t even been thought of yet. But it may be decades before fully autonomous vehicles are widely adopted, if ever.
Rumors of the demise of Americans visceral enjoyment of the act of driving, like sex, are greatly overstated and premature.