SCIENCE AND TECH
In Depth: Self-Driving Cars are Coming. Are We Ready?
With carmakers racing to get self-driving models to market by 2020, will cities, governments and people be prepared? Here's what the experts think.
Two thousand and twenty. According to Toyota we will have self-driving cars – or at any rate, something very close to them – by 2020, just a few short years from now. When last October the automaker unveiled its plans to start selling a sensor-packed, super-smart vehicle within five years, it was a declaration that made the prospect of cars that do all the work for you much more than just a sci-fi fantasy. Now it’s a promise.
What made this announcement so significant was that much of the attention around self-driving vehicles had been paid to the automotive outliers, the innovation brain trust such as Google, Apple and Tesla, and their high-profile efforts to rapidly define the future of human mobility. But with mass-market players like Toyota, GM, Ford and many others – along with the luxury brands like Mercedes, Audi and Lexus – rapidly driving forward to bring us cars that drive themselves, the potential for the technology to be adopted en masse becomes much more likely.
The innovation being rolled out soon – and even now, as we speak – is making the need to address autonomous vehicles from a policy perspective all the more urgent. The carmakers are, in effect, forcing the hands of legislators to write new laws, allow the testing of new hands-free vehicles, and generally make sure the public is ready for a world without drivers.
A lot needs to happen in four short years before this Jetsons-like dream can become reality. The pressing question then becomes, how ready are our cities to actually meet that brave new future on such an accelerated timeline? The answer, not surprisingly, is complicated, but in short you could say: The road ahead is long, but we will arrive at our destination.
AP Photo/Jae C. Hong, File
THE FIRST BIG QUESTION: WHY GO DRIVERLESS?
Before we dig deep into the question of whether we’re ready, we ought to understand the end goal for driverless cars. Why do we want them in the first place?
For consumers, the simple allure of self-driving cars may be the ability to kick back and have a machine drive you to work. For the experts, the the potential is utterly transformational, though the most exciting aspects depend on whom you ask. Some breathlessly go on about how not having to focus on the road will boost productivity, or passive time while commuting. If Volvo and Netflix have their way, you’ll be watching movies while in the erstwhile drivers seat. One expert even predicts that people will have more sex behind the wheel.
Many look to the collective benefits, such as radically transforming mass transit and cities. Other people identify the significant benefit to the elderly and immobile. Others still rhapsodize about how a car that can drive without human input can become a source of income and take the sharing economy to the next level, thus removing the need for more cars on the road – in fact, in its Disrupted Mobility report Barclay’s predicts that U.S. automobile sales will decrease by 40% by 2025.
THE HOLY GRAIL? ELIMINATE HUMAN ERROR
Still, all of those benefits are secondary to the main outcome sought by experts: safety.
“For people like me working in this industry, (having no drivers in the car) is literally our Holy Grail,” says Raj Rajkumar, professor in the department of electrical and computer engineering and co-director of the General Motors-Carnegie Mellon Vehicular Information Technology Collaborative Research Lab. “We want to get to a point in time where humans, which are literally the source of accidents, can be taken out of the equation."
Rajkumar – who also holds a courtesy appointment in the Robotics Institute at CMU in Pittsburgh, the birthplace of autonomous vehicle technology – says automotive accidents cause 1.25 million deaths a year globally, and humans are usually to blame.
“We want to get to a point in time where humans, which are literally the source of accidents, can be taken out of the equation."
“That’s a lot of people dying,” he says. “If that many people die, the number who get injured is at least 10 times that figure and the number of accidents about 30 times more than that figure. And more than 90% of those accidents are due to human error or distraction.” Computers, on the other hand, are not prone to the same problems, so the hope is that when computers do take over, accidents, injuries and fatalities will lower dramatically. “So the promise and potential is enormous.”
SAVING LIVES AND MONEY
When translated into economic terms, it’s staggering. Ravi Shanker, Morgan Stanley lead analyst covering the North American auto and related industries, wrote a research paper on the topic: “Beyond the practical benefits, autonomous cars could contribute $1.3 trillion in annual savings to the U.S. economy alone, with global savings estimated at over $5.6 trillion.” (All figures in U.S. dollars.) That translates into 8% in annual GDP, the reallocation of which could have major implications of its own, according to the report.
In Canada, a joint report issued in January 2015 by the Conference Board of Canada, the not-for-profit Canadian Automated Vehicles Centre of Excellence and the Van Horne Institute showed that the cumulative effect (including collision avoidance, fuel cost savings, and congestion avoidance) of AVs will benefit the Canadian economy to the tune of $65 billion per year, or 3.6% of GDP.
Aside from cost savings, the predicted size of the autonomous vehicle market alone is impressive. According to a report from Boston Consulting Group, the market could reach $42 billion by 2025, and autonomous cars could make up a quarter of global auto sales by 2035.
AP Photo/Tony Avelar, File
WHERE ARE WE NOW?
In order for cars to be completely uncoupled from a human hand, significant progress needs to be made in a number of areas including technology, government policy, security, insurance and human psychology, and we still have a lot of thinking to do on each of these points. It’s the legislative hurdle that’s arguably the most difficult, however – or at least the factor that could respond the slowest to change.
That change is really happening, meanwhile: Driverless vehicles are already being used in industrial settings, such as Suncor Energy’s fleet of 175 autonomous trucks that are being used in its Alberta mining operations. Among consumer vehicle manufacturers, Tesla is working on a semi-autonomous autopilot system. So is Volvo, which recently took a chance to pooh-pooh Tesla’s efforts. Audi expects its A8 to be equipped with a piloted driving autonomous system by 2017. And Google is testing completely driverless vehicles designed to shuttle people around; to date it has logged 2.3 million kilometres (1.4 million miles) on public roads, the most of any carmaker.
While Google’s little pod-like vehicles paint a driverless vision of the future (a slightly twee one), most technologies being tested by carmakers are tools to assist a driver, not replace one. For example, Toyota refers to the technology it recently demonstrated as a “mobility teammate concept,” meaning that while it significantly aids in driving, there’s still a steering wheel, and a human being is still required.
In the United States, it’s up to individual state legislators to determine whether or not to permit self-driving cars on their roads. In 2015, 16 states introduced legislation related to AVs, and seven states currently allow road tests. California, Nevada, Michigan, Florida and Pennsylvania along with the District of Columbia have been among the first to welcome the innovation but the number is likely to grow rapidly.
"In 2016, we are going to do everything we can to promote safe, smart and sustainable vehicles. We are bullish on autonomous vehicles.”
In January 2016, the U.S. Department of Transportation announced that it was developing policies for self-driving cars in a bid to get the vehicles on the road more quickly. It might even waive regulations to do it, according to U.S. Department of Transportation (DOT) secretary Anthony Foxx. He said that the National Highway Traffic Safety Administration was going to spend six months developing guidance for automakers, and develop a model policy for states to follow if they decide to allow autonomous cars on public roads.
“In 2016, we are going to do everything we can to promote safe, smart and sustainable vehicles. We are bullish on autonomous vehicles,” Foxx said during an appearance at the North American International Auto Show in Detroit.
But the bullishness doesn’t stop there. Foxx also announced that the government’s budget would provide $4 billion over the next decade for programs to test connected vehicles, including the DOT Smart City Challenge, a $50 million competition to remake the American city. Seven finalists were announced (two more than expected as the DOT, says Foxx, was “blown away” by the submissions), and a winner will be revealed on June 1.
Jesse Berst, founder and chairman of the Smart Cities Council, a city-building coalition, says this challenge is a boon to the AV movement. “One of the things that makes me confident is that this DOT challenge grant specifies autonomous vehicles as one of the things they really most want [addressed]. As I'm travelling around the U.S., so many cities are working on grant proposals to try and become the winning city. Whether or not they win, they've got a plan now because they had to build one in order to apply. So I think that may be a bit of a forcing function that causes some experiments to show up a little sooner than they might have otherwise.”
In Canada, Ontario is the only province to have officially addressed the prospect of driverless vehicles. In October 2015, the Ontario Department of Transportation announced that starting January 2016, companies could apply for road tests, which are expected to begin in March. For automakers, Ontario (along with a states like Michigan, where the University of Michigan launched MCity, a replica city designed for connected and automated vehicle testing, and where Toyota is planning a $1 billion AV research hub),provides something that test environments in sunny states can’t offer: snowy, winter roads, which pose an important and difficult technological challenge.
Transportation Minister Steven Del Duca says coming up with a plan for autonomous vehicles was one of his first mandates when taking that post in 2014. The goal: to be a leader in Canada.
“Ontario is one of the world's – certainly one of North America's – leading jurisdictions as it relates to the automotive sector. So, from the standpoint of economic development potential, we wanted to make sure that Ontario emerged as a leader so that when companies are looking for locations in which they can invest, they can create jobs, they can innovate, they'll see Ontario as a jurisdiction that embraces the technology, embraces the innovation, and they'll see a government that wants to partner with them on this,” says Del Duca.
GREAT DISRUPTION, GREAT OPPORTUNITY
Of course, bringing this brave new world into being will not come without costs, both economic and human. Whenever a change is billed as disruptive and transformative, it’s guaranteed that there will be winners and losers – at least in the short term.
Among those (and this is nowhere near exhaustive) that face disruption: the automotive industry, because the components required for high-tech vehicles will require new skillsets and the purchasing cycle of vehicles will change; the insurance industry, as liability considerations are a looming issue and rates – and revenues – stand to fall along with human-error accident rates; and those who drive for a living, who are likely to be a significant casualty of this innovation. By some measures, "truck driver" is the most common occupation in nearly every US state, and in Canada it's the second most common job for men.
There are even impacts that reveal themselves only when we think more broadly and abstractly about the butterfly effects of driverless cars. Even a drop in road fatalities isn’t without unintended effects. Barrie Kirk, executive director of CAVCOE (who also envisions a spike in vehicular coitus) sees a downside to fewer vehicular deaths. “What happens when healthy young people, who are most likely to die in car accidents, stop dying? What’s the potential impact on people relying on organ donations,” he muses.
Still, given the potential opportunities, Kirk applauds the Ontario Ministry of Transportation for being quick to invite road testing, and for making the process simple to understand, with a one-page application and regulation document. But he says when it comes to being a worldwide leader the province is “in general doing quite poorly.”
“A lot of companies could be involved but they’re not,” says Kirk, citing specific opportunities for the automotive parts industry to prepare itself for a pivot to new components, and for the technology industry to create the necessary software for these smart, connected vehicles. This ranges from components like sensors and cameras all the way to encryption software to secure AVs from malicious hacking.
That said, Kirk points to a few Canadian stars that are leading the way. Blackberry subsidiary QNX has developed an operating system that is used in 60 million cars, or over 60%, worldwide, Kirk says. And, after years of investing in infrastructure such as citywide WiFi, the southwestern Ontario city of Stratford has positioned itself as a connected city in a bid to become a real-world destination for autonomous vehicle testing. CAVCOE also recently released an unsolicited white paper outlining 30 recommendations for the Canadian federal government to help stimulate this burgeoning industry.
For his part, Del Duca says his government intends to enable positive outcomes. “We believe there's a huge potential from an economic standpoint here in Ontario to attract jobs that come as a result of this innovation. We want to embrace innovation.”
THE FIRST STEP: FOCUS ON THE PUBLIC
While in Stratford the vision is to create a city where cars communicate with infrastructure and each other, Jesse Berst of Smart Cities Council sees another great opportunity that’s perhaps even more readily attainable: to revolutionize mass transit.
With self-driving vehicles matched to fixed routes, Berst sees an opportunity to make transit more efficient by servicing communities that need it without massive infrastructure expenditures. As new neighbourhoods develop with more housing and business, new routes, perhaps with smaller shuttle vehicles, could easily be added, creating a flexible and responsive structure, he says.
“That could really solve this problem we see everywhere, which is: the people who work in the city at some of the various central service jobs often can't afford to live anywhere near work.”
He also sees AVs having a revolutionary effect on city building. “Most cars sit idle about 95% of the time,” he says. “But what if I could be in my office and my vehicle could be helping others? I could be in some kind of a car-sharing program, so you have fewer cars because the ones you have will be utilized to a much greater degree.” With cars in a constant state of use, the need for parking – which Berst estimates takes up a quarter to a third of downtown real estate – suddenly frees up.
HOW DO WE GET THERE? SLOWLY, WITH ONE HAND ON THE WHEEL
So how likely are we to really see self-driving cars on the road by 2020? Not very, say experts. The consensus is that we’ll see significant advances in autonomous technology, while full-out driverless vehicles will have to wait.
Carnegie Mellon’s Raj Rajkumar sees a good sign in the fact that all the major carmakers and leading information technology companies are active in the space, and that they’re investing what he estimates is billions of dollars. “With so many smart people are spending time on [AVs] and once you start investing those kinds of resources, clearly progress will be made. That being said, let's be cautious about how long it's going to take to solve some of the underlying problems. Lots of progress is being made but it’s still a long road ahead.”
Rajkumar predicts that we’ll see a more developed version of the incremental deployment path that we’re already on, pointing to Tesla’s recently released automated highway feature, which will be available on its hotly anticipated Model 3, and GM’s intention to have an automated highway feature in the Cadillac global brand in 2017. And automakers like Chrysler and Ford are forging partnerships with tech players like Google and Uber to make sure they’re on the right side of history. “Pretty much every carmaker has a prototype along those lines, so incremental driver-assist features where the vehicle is driving itself, but on only particular segments of the road with a human in the driver's seat is going to happening on a regular basis moving forward.”
Berst sees first rollouts of autonomous vehicles in controlled environments, such as on bus routes and HOV lanes. “I think we may see some great pilots and experiments by 2020.” Those pilots, he says, will likely be more fleet oriented, like transportation in airports and industrial parks (look to companies like France’s Navya, which is road-testing its autonomous shuttles in Switzerland, or WePod, which is undergoing trials in the Netherlands, as models for transportation). As for driverless cars roaming the streets? “That may be more like 2025 or even 2030,” he says.
"If we remove drivers prematurely, it will end up causing an unsafe situation, which is bad for everybody involved."
As it turns out, a protracted rollout might be necessary for the most unpredictable hurdle of all: human perception. An undercurrent bubbling below all of the excitement around self-driving cars is how people will actually feel about completely relinquishing personal agency to such a new technology.
“I think for a lot of people who are used to participating in their own transportation … part of the challenge is cultural. Part of it is people getting over that mental hurdle of “is it safe?” says Del Duca, who sees self-driving cars about seven to 15 years away.
That question has gained media attention, too. Bloomberg recently wrote a piece that questioned whether billions are being spent on a technology that nobody wants, citing a statistic that just 23% of Baby Boomers trust self-driving cars. Yet the same report from J.D. Power says more than 55% of people born since 1965 are already comfortable with driving-not-driving.
What we really need is time to get used to the idea, says Carnegie Mellon’s Raj Rajkumar, referencing the deployment of cruise control as a cautionary tale. While it was revolutionary in the late 1980s of cruise control, it’s a comparatively simple technology yet it still took several years for the problems to be ironed out.
“If technology causes a problem, then it’s not clear how society will react and there could possibly be a backlash. If we remove drivers prematurely, it will end up causing an unsafe situation, which is bad for everybody involved; the public as well as the carmakers. So I do think it will take about 10 years before we see truly driverless vehicles on the road.”
With files from The Associated Press