Driverless cars are an alluring idea, and a handful of companies have made amazing technological progress. This has led to some pretty bold predictions of driverless cars by 2025 or sooner. I believe it will take a quite a bit longer due to technological, cost, consumer behavior, regulatory, and legal barriers that will slow down development and adoption.
Benefits of driverless cars
Driverless cars could generate tremendous benefits for our society -- saving lives and billions of dollars. Most accidents are caused by human error. Over 30,000 people die each year in traffic accidents in the United States, roughly 40% of which are caused by alcohol, distraction, drugs, or fatigue. Driverless cars hold the promise to greatly reduce accidents, savings lives and billions of dollars, to improve fuel economy, alleviate congestion, and a host of other benefits. It's an exciting idea, and there's been some pretty amazing progress in the past few years.
Recent progress in driverless cars
Google's (GOOGL 0.55%) self-driving car project has been widely publicized. Google employees are now testing the driverless cars on the freeways in Silicon Valley, and reportedly the test vehicles have driven more than 500,000 miles without a single accident. Last year, Sergey Brin boldly announced that "you can count on one hand the number of years" until ordinary people will experience driverless cars, implying a 2018 release.
Elon Musk announced on Twitter that Tesla (TSLA 12.06%) is working on a "practical autopilot system" for the Model S. Apparently, he's forming an autonomous driving team, which will report directly to him. And, Musk, who doesn't back away from challenging projects or bold statements, told the Financial Times in September that Tesla will have a 90% autonomous vehicle within three years. The fully autonomous model will come soon after that.
The big auto-makers are also pushing toward driverless cars. Volvo plans pilot driverless cars on the roads of Gothenburg, Sweden in 2017. Nissan recently announced that it would sell driverless cars by 2020. According to the Economist, BMW (BAMXF 0.05%) has been testing driverless cars in Munich.
Hype and unrealistic expectations
The potential benefits of driverless cars are very alluring, and recent pilot programs by Google and others are impressive. Add plenty of media coverage and bold pronouncements from people like Sergey Brin, Elon Musk, and Carlos Ghosn, and you've got a recipe for hype. Driverless cars are certainly the future. Raj Rajkumar, a Carnegie Mellon professor, recently told Congress, "It's not a question of if, but when" autonomous cars debut.
That's probably true, but driverless car promoters are ignoring many of the challenges associated with development and adoption. And some investors are getting sucked in. Case in point, Steve Kuhn of Pine River Capital recently appeared on Bloomberg TV to explain his short-case for Progressive (PGR -0.85%). He believes (and has bet his hedge fund's money) on the fact that driverless cars will eliminate the need for car insurance in the near future.
In my opinion, driverless cars are going to be slower in coming than predicted by Kuhn, Brin, Musk, or Ghoson. The development of driverless cars still faces big technological hurdles, which could take decades longer than expected to solve.
And technological barriers are only half the battle. A host of other challenges, such as cost, consumer perception, regulation, and liability concerns, are likely to slow down the spread of truly driverless cars.
Technological hurdles
The most obvious hurdle to driverless cars is technological. While there have been tremendous strides in automated driving sensors and software, replacing a human is a hugely difficult task. Things like rotaries, freeway merging, detours, and pedestrian-avoidance haven't been fully mastered. This is why Google's driverless cars are backed up by trained test-drivers that frequently need to put their hands on the wheel.
According to estimates published in Popular Science, even the best pedestrian-avoidance systems will hit a child running into the street about 1% to 2% of the time. According to Brad Templeton, an automotive consultant and former member of Google's self-driving project: "Obviously, 99% just isn't good enough; we need 99.99999, and what people don't seem to realize is that the difference between those two numbers is huge. It's not a 1% difference—it's an orders of magnitude difference." To improve the technology by an order of magnitude could take decades longer than optimists expect.
Cost barriers
Even if driverless car technology is perfected, it will likely be cost prohibitive for many years. Today, the cost of automated driving equipment and software ranges from $70,000 to $100,000, which is in addition to the cost of the car. That's far beyond the means of most Americans -- last year, the top 27 selling vehicles in American ranged from $16,000 to $27,000. Of course, the costs are expected to drop, but it could take many years for it to be low enough to spur mass adoption. In a recent J.D. Power survey, only 20% of consumers say they would pay an additional $3,000 for a driverless car.
Consumer perception
Obviously, the idea of enjoying a latte and a crossword puzzle while your car drives sounds pretty cool. But driverless car enthusiasts may be overestimating how rapidly consumers will embrace the technology. New technologies require new consumer behavior, and consumers don't always adopt them quickly as we forecast, despite all the perceived benefits.
Additionally, the idea of a driverless car challenges a key human psychological bias: illusion of control. People feel safer if they're "in control," i.e. driving the car. For many years, statistics have shown that flying is safer than driving, but fear of flying certainly exceeds fear of driving. A driverless car, which requires surrendering control, may induce more irrational fear than expected.
Legal and regulatory impairments
Three states (Nevada, California, and Florida) and the District of Columbia have explicitly legalized driverless cars. In the rest of the country, it's not clear whether driverless cars are legal.
According to the leading expert on the subject, Bryant Walker Smith of Stanford Law School, driverless cars are "probably" legal in the United States. The National Highway Transportation Safety Administration (NHTSA) is still only in the initial research phase of establishing standards for driverless cars. This initial research phase by NHTSA is expected to last up to four years. At present, the NHTSA doesn't recommend allowing driverless cars on the roads, except for testing.
So we're not even close to national standards, and it's likely to be a contentious political issue. Driverless car accidents, which are inevitable, will likely receive widespread publicity. It could take decades of political fighting to overcome doubts about the safety of driverless cars and pass workable, national legislation.
Liability confusion
Liability concerns could also prove a major roadblock. Who is responsible for a driverless car crash -- the driver, the auto manufacturer, another third party? It could be a litigious nightmare.
Even Chris Urmson, the director of Google's self-driving-car project, admits this is a problem. According to him, "What's going to happen, no matter what the law says, is people are going to get sued." Of course, he also points out that there wasn't legal protection for the Wright brothers, but that didn't stop aviation. And while legal and liability concerns probably won't stop the development of driverless cars, it will certainly slow down development and adoption.
Foolish takeaway
Don't get me wrong, I'm excited about the idea of driverless cars. No doubt, it will happen in the future, but it will probably take quite a while longer than some have suggested. I'm a very long-term investor (willing to hold for a decade or longer), but I wouldn't make an investment based on the expectation of driverless cars in the near future. It just seems too far out. Of course, the future is uncertain, but there seem to be too many stumbling blocks for wide-spread adoption of driverless cars anytime with the next 20 to 30 years.
