Wait, Who Even Asked for Self-Driving Cars in the First Place?
Five years ago, autonomy wasn’t even a thing. Now it’s the only thing. Why?
The path to the place the automotive world is now—embroiled in breathless discussions about a new technology that everyone in the world needs to prepare for right now—actually began in another industry's yard. Self-driving cars didn’t spring from any automotive minds, as some might expect, but rather, from a military yearning for battlefield efficiency. It began, for all intents and purposes, with the DARPA Grand Challenge, a 150-mile desert race announced by the Pentagon’s military-technology R&D agency in 2002 and executed in 2004.
Not a single vehicle finished that first race, but in the same event a year later, five did. That staggering leap forward in autonomous-driving capability in just one year—in an off-road race, no less—showed how fast this technology has moved. It also shows exactly where the seeds for our current infatuation were sown. That is, not at Ford, Honda, Volkswagen, General Motors, Mercedes-Benz, or Audi—but rather at Stanford, Carnegie Mellon, NC State, Ohio State, and a slew of other schools and technology firms populated with robotics engineers and software geniuses all gravitating toward Silicon Valley, to be snapped up by Google, Apple, Uber, and others. To the extent that there was an auto-industry presence in the original Grand Challenges, it was as primarily in providing the mules for these brainiacs to install their hardware and software.
But that doesn’t entirely explain the current industry-wide obsession with autonomy, given that those events were held roughly 15 years ago. Just five years ago, nobody in the auto industry—let alone the public—was really speaking about self-driving cars as a practical near-future reality at all. Nowadays, it’s all you hear about (right alongside electric cars); with Google all over the technology and Apple possibly so, fantastic driverless concept cars peppering auto shows, and Teslas and Cadillacs ably chauffeuring us down highways...or with Uber mowing down pedestrians in Arizona, depending on your interpretation of our current state of progress.
So what brought on the current surge in interest in autonomous driving, and the seemingly frantic, if not desperate, pace of development? Technological readiness? That’s a part of it, sure—though it still has extremely far to go in terms of truly being able to navigate all driving conditions, regardless of how impressive highway driving might seem. Government mandate? Not at all. While military interest originally spurred the technology on in the early Aughts, the government has been reactionary—if you could even call it that—in its urging of autonomous technology, despite the massive safety benefits it could ultimately bring. Consumer interest? Another big no, at least broadly speaking. As is the case with electric cars, consumers haven’t actually asked for autonomy directly, and while studies show that consumers appreciate the safety benefits early semi-autonomous features provide, they’ve yet to prove that consumers are all that excited about turning over the act of driving to a machine. Seems some people actually like driving, if you can imagine. (Or at least they're wary of computer drivers.)
That essentially leaves the car manufacturers and the tech industry as the ones foisting this upon us, actual demand be damned—and indeed, the race to autonomy currently looks a lot like a very recent case of very explicit one-upmanship between these two industries. Or more pointedly, a panicked, full-stop realization by the carmakers that the tech industry was about to totally eat its lunch. So in spite of the fact that the auto industry has historically shown little interest in taking driving away from its customers, and in spite of the fact that there’s no definitive interest in or demand for autonomous technology—remembering, of course, that benefits, safety, convenience or otherwise, don’t always equate to demand—every major manufacturer has cozied up to autonomy startups, opened their own “mobility” research and development centers in Silicon Valley, and aggressively announced their hot pursuit of autonomous driving technology.
So that means, of course, that it’s entirely Silicon Valley’s fault. Sort of.
“The answer is multidimensional,” said autonomy engineer Missy Cummings, director of the Humans and Autonomy Laboratory at Duke University. “Google has actually been at this a long time. I rode in one of their cars in 2012, and the Grand Challenges happened well before that. But I do think we hit a tipping point in approximately 2015 due to many factors. Elon Musk was gaining steam [with Tesla], commercial drones were becoming hot, and in general there was a tangible shift in the hype surrounding anything that has to do with artificial intelligence. In the World Economic Forum in the past 3 years, AI has gone from a curiosity to a centerpiece of conversation.”
Cummings goes on to argue that we’re in a bubble, particularly around artificial intelligence, the key tech enabling behind autonomy. “While there is certainly a lot of great potential for AI, it is generally grossly misunderstood and overrated across the business sector,” she said. “People want to believe it’s a magic bullet but the reality is it is just another tool in the tech toolbox and not the panacea they are hoping for.”
But the catch-up mentality is also a key factor in the current frenzy, according to Tasha Keeney, a researcher and autonomy analyst at ARK Invest. “After the DARPA Challenges, technology players, namely Google, started working on autonomous tech and were really the trailblazers,” she said. “Daimler and Volvo had been working on semi-autonomous features too for some time, but were moving very slowly. For a while most OEMs were saying that autonomous cars would never happen, or would emphasize how many years and years away it was, or how dangerous autonomous cars could be.”
She found this unsurprising, given the fact that traditional OEMs really weren’t equipped to develop autonomous cars, lacking both the software expertise and the talent to push the technology along. “If you think about it, it was really in the OEMs best interest to scare the public away from autonomy because they were unable to tackle those projects themselves,” Keeney said. “Then, the OEMs got scared. They realized autonomous cars were coming whether they participated in that movement or not.”
Ultimately, the impact of all this fast-track development could fully justify the automakers' concerns. ATK thinks the auto industry will experience “severe consolidation” over the next decade, and if the carmakers don’t have successful electric and autonomous platforms, many of them will not succeed. Throw in the fact that scrappy startups are digging in with solid results, and you have a perfect layup for full-scale industry transformation.
“Tesla contributed to those fears because it emerged as a successful startup in an industry that typical squashes startups,” Keeney said. “Tesla was built from the ground up as a software powerhouse and it's uninhibited by the dealer network, so it’s much more nimble than traditional OEMs—and so are technology companies like Google. The typical design cycle for an automaker is fibe years, but Tesla is shrinking that to one to two years."
As for whether the consumers will come around to the technology and eventually spur demand along on their own, it seems likely. A study published in January by Deloitte found that less than half (47 percent) of U.S. consumers felt that autonomous technology was unsafe—which the company noted was a dramatic decrease from 2017, when 74 percent felt autonomous vehicles would not be safe. This study, of course, predated last month’s fatal Uber crash in Arizona, but events like that tend to generate spikes in concern rather than trends—at least, until a pattern of similar incidents emerges. But if the successes build the technology proves itself, and everyone’s lives become safer and less stressful, the pendulum could swing from being “not afraid” of it to “desperately wanting it.”
"The market for autonomous cars will be massive—$10 trillion by the early 2030s, and platform providers like Google or Tesla could take a 20 percent of that," Keeney predicted. "This alone could entice market entrants.”
Keeney also thinks consumers will be motivated by the economic impact of autonomous technology, further sealing the deal. “Price will be pretty compelling, with autonomous taxi services costing 35 cents per mile—or roughly half the cost of driving a personal car today and 1/10th the cost of a taxi,” she said. “I don’t pay too close attention to consumer surveys because I don’t think the typical consumer is recognizing all the benefits and the full picture when they log their responses.”
So in the end, what began as a dusty desert race dominated by scrappy, tech-savvy innovators could very well be won by...well, those same scrappy, tech-savvy innovators. They'll just likely be far richer by the end of it all.