This article was published 22/8/2015 (2510 days ago), so information in it may no longer be current.
It’s 7 a.m. on a sunny Monday morning in the not-too-distant future, like a hundred Mondays before. You eat breakfast and hop in your car. As the garage door opens, sonar, cameras and radar guide you out and determine if it’s safe to exit the driveway.
Forty-five minutes later, you’re at work; you walk in and begin your workday.
What’s different? You didn’t drive. Your car did.
Driverless cars, where human driving skills are replaced by those of a microchip, are coming sooner than you think — possibly before this decade is out.
Nissan, which has vowed to have self-driving cars on the road by 2020, and Volvo have already tested such cars on public roads, proving them adept at such tasks as cruising, merging, using an off-ramp and avoiding an obstacle. Google has a fleet of self-driving cars.
Other carmakers are following suit, and the seeds of the technology are already available to the public in cars such as the Mercedes-Benz S-Class or Acura MDX, which will take over for you in stop-and-go traffic. These cars will brake, accelerate and steer up to a certain speed, when you take over.
Already, in mines and some college campuses, driverless vehicles are hauling ore to the surface or students to class. On some Manitoba farms, tractors are out working fields. By themselves.
The implications of driverless cars are profound, and promise to reshape the world we live in. Cities may look vastly different. Downtowns will be freed of the need for immense investments in parking. Surface lots, considered the bane of downtown urban planning, may well become unprofitable to the point owners might finally have incentive to develop buildings.
A suburban office building will no longer need to be an island rising out of a sea of parking spaces, perhaps rewriting the business case to locate in suburbs entirely. Absent the need to provide acres of parking spaces, companies may decide to move back to downtown areas.
More people may choose to live farther away from work and more people will be induced into using public transit.Consider: No longer will your car be tied up at work all day. You can have it drive you to work and then tell it to return home. When it arrives, your spouse or child has a car for the day. At 5 p.m., it’s back, waiting for you outside your workplace. Instead of two cars, your family can get by with one, if you own a car at all.
There will be no more "park and ride." Public-transit users who currently drive to a transit hub and park will be able to simply send their car home. Where parking is required, much less space will be needed for the same number of cars, which will be freed of needing to leave space for the occupants to get out. The car can drop you at your destination and then park somewhere with only inches to spare. With the prospect of vehicles that communicate with each other, it may not even need to leave space for itself to get out: It can simply tell the cars ahead or behind to move when needed.
Valet parking and searching for a parking spot may become a thing of the past, as you can use what some futurists humourously refer to as f***-off mode, where you arrive at a destination, get out and tell your car to... well, you get the idea. It goes away, finds a safe spot to hang and comes back when summoned.
These are just some of the predictions in a few reports this year from highly regarded researchers, including the Conference Board of Canada and the University of Michigan Transport Research Institute.
The Conference Board warns governments to start planning now for what it calls Automated Vehicles (AVs), with new laws, new urban-planning strategies and new ideas for roadway construction.
The board’s report is entitled Automated Vehicles: The Coming of the Next Disruptive Technology. It paints a fascinating, and, at times, frightening picture of a future without drivers.
The 72-page report lays out some benefits — reduced collision costs, savings in fuel costs, increased productivity — and some negative implications — increased urban sprawl, job losses and infrastructure costs to accommodate automated vehicles.
"AVs will be nothing less than the first widely available ‘autonomous robots’ to be used by nearly everyone in the world’s advanced economies," the report states.
Perhaps the most interesting part of the report — much of the technology has been demonstrated already — was the predicted impact on urban planning.
With nothing to do on the way to the office, the board suggests many vehicle occupants will use the commuting time to work — catching up on emails, finishing reports, making phone calls — all of which isn’t currently legal or advisable.
Because commuting time will suddenly become productive, the report predicts many people will be encouraged to seek out lower-cost housing further from city centres. That will put a major crimp in any effort to contain urban sprawl, creating potential infrastructure headaches for cities, some of which are already straining to service outlying areas.
"There is a strong likelihood that people will be willing to tolerate longer commutes if they are able to be productive in the vehicle, especially if it means that they can buy cheaper housing as result of the longer commute. This will result in sprawl, and a possible reduction in land values in existing suburban and ex-urban areas," the report states.
Robert Galston is a Winnipeg master’s student in urban planning and a frequent commenter on urban-planning issues. While he sees many benefits to automated vehicles, he’s not convinced a big drawback will be urban sprawl.
"The ability to browse Twitter or play Candy Crush a little bit longer might not outweigh the benefits of being physically closer to things," he said. "Technological improvements have not lessened the need for face-to-face contact in the knowledge economy.
"Being able to get to work earlier to meet with colleagues or clients over breakfast, for example, is more important than being able to email them from the road."
Still, Galston sees an upside. "These kinds of technologies and efficiencies work well for mid-sized cities such as Winnipeg, which are not dense enough to support subways or an endless supply of taxicabs, but are big enough to have demand for alternatives to car ownership, which growing numbers of millennials and retiring baby boomers are looking for."
Car sharing may well become the next boom business, possibly even supplanting traditional taxi, car-rental services or even automotive sales. Since an AV doesn’t need to be parked, one customer could use it to get to his destination, then it could go on to pick up another client, overlapping customers and rarely being out of service. As well, a fleet of AVs could be managed centrally so customers know an available vehicle is never far away.
The University of Michigan has built a mock city to test driverless vehicles. A recent survey by the university indicates only 15 per cent of respondents want to turn over driving completely. (University of Michigan)
The ability to overlap multiple trips, particularly among family members, may drastically reduce the demand for vehicles, write Michael Sivak and Brandon Schoettle, of the University of Michigan Transport Research Institute, in their report Potential Impact of Self-driving Vehicles on Household Vehicle Demand and Usage.
Their analysis was comprehensive and complicated. Suffice it to say, they wrote a computer algorithm to predict the impact of AVs on vehicle usage based on U.S. driving statistics. They showed demand for vehicles per household being cut nearly in half, while the miles driven per vehicle yearly nearly doubling.
John Estrada, CEO of advocacy group and trade journal Driverless Transportation, said some estimates have shown that by overlapping clients, the current number of vehicle miles driven could be covered with 80 per cent fewer vehicles.
The news isn’t all bad for carmakers, however. Sivak and Schoettle point out doubling the miles driven per vehicle will increase the repair and maintenance business and reduce the lifespan of vehicles, with the resulting turnover at least partially offsetting the decline in demand.
Automated vehicles also have the potential to completely change the automotive business landscape. Jac Nasser, who was president of Ford Motor Co. from 1998 to 2001, touted a plan where instead of buying a vehicle, you’d contract with the carmaker for a certain amount of personal mobility. The advantage, he said, was you could access that mobility wherever Ford operated. Taking a trip to Britain? Use your contract to drive a Ford in that country. Needs changed? Alter your contract so instead of, say, a four-door sedan, you now can drive an SUV.
Think of it as time-sharing, but for cars.
With automated vehicles, Nasser’s plan might finally make some sense.
The Conference Board report lays out several benefits to AVs, including reduced fuel costs, since AVs will use one predicted bit of infrastructure called smart roadways, where traffic authorities share real-time information about construction, congestion or other delays with cars, allowing the cars to plot different routes to save time and fuel.
Estrada said with cars talking to each other, you could stuff them into a special traffic lane at much higher numbers, since they can be spaced much more closely together. (By talking to each other, when braking, for instance, the lead vehicle isn’t merely applying its own brakes, it’s applying the brakes of every vehicle in the convoy.)
Galston said the technology will alleviate some of the human factors that create congestion, such as driver inexperience or lack of training, but doesn’t overcome the fact demand always seems to expand to fill available road space, particularly while driverless cars are still sharing road space with driven vehicles.
A big question, particularly in cities such as Winnipeg, is how will driverless vehicles respond to snow? Slippery roads are easily handled by technology, but many of the systems in use today require clear lane markings to set the position of the vehicle. What happens if the markings are under a windrow? Will a driverless car know what to do about ruts?
The Conference Board also predicts a rise in the use of public transit, as the people now beyond the reach of public transit who opt not to park and ride choose to do so, particularly since the "park" in park and ride is eliminated.
"Similarly, the need for large, expensive transit park-and-ride lots will decrease," the report says. "People will be able to travel from home to a transit station via AV and then send the AV home, possibly to enable other family members to go to work or school."
For transit operators, such a system will be a boon, the board suggests, as the transit system would only ever need to service hubs.
"This will significantly reduce the cost of the ‘last mile’ portion of using public transit."
The Board predicts massive job losses as delivery drivers and taxi and limo drivers are rendered moot. As well, the report suggests an unpredictable impact on the hotel industry, as some drivers who visit nearby cities for work may opt to replace overnight hotel stays with having their cars drive them home.
Quoting a report by U.S. investment firm Morgan Stanley, the report suggests worldwide savings of $5.6 trillion per year, primarily from reduced collision-repair costs and productivity gains.
Those productivity gains come in a few ways: people can work while on the road and when they get where they’re going don’t need to spend a nanosecond looking for parking, just get out and let the car take care of itself.
There are also expected to be savings in fuel consumption, as congestion is avoided and acceleration, deceleration and cruising are all programmed for optimal fuel economy.
Of course, those savings come at a cost: the Morgan Stanley report predicted US$488 billion coming out of the American collision-repair industry. That means lost jobs.
Most reports on AVs read as though they were prepared by people who either don’t drive or don’t enjoy it when they do. Perhaps it’s merely the dispassion of the scientific method, but there isn’t a single passage in either that even hints at lamenting the loss of an activity many have held dear since the dawn of the automobile age.
Steven Polzin, director, mobility policy research, at the Center for Urban Transportation Research at the University of South Florida, said people who love to drive don’t have much to worry about.
But he did say automated vehicles do have the promise of giving the freedom of mobility to people who otherwise can’t drive. From elderly people to people with mobility issues to the young, AVs will one day be able to take them places only taxis, Handi-Transit or willing parents can now.
Getting there requires resolving several issues, issues into which research is only now beginning: liability is the key one. In other words, if two driverless vehicles collide, who is at fault? The vehicle owner? The carmaker? Whose insurance pays to repair the damage?
Polzin said the issue of liability will be a sticking point, and said some of the ideas coming out now include changing the insurance model or even rewriting the business plan of carmakers where ownership of automated vehicles remains with the carmaker, in a lease or sharing arrangement. "It’s really premature to speculate."
Driverless Transportation’s Estrada was ready to speculate, however.
"At some point, the carmakers are going to have to sign on for the liability," he said. If both parties to a crash are in vehicles supplied by a carmaker with full autonomy "you can’t really say one is at fault and the other is not."
Another key obstacle, according to a subsequent report by the University of Michigan’s Brandon Schoettle and Michael Sivak, is overcoming staunch opposition to self-driving cars from today’s drivers.
Their report, Motorists’ Preferences for Self-Driving Vehicles, found most drivers overwhelmingly favoured either zero self-driving ability or only partial self-driving ability in their cars. Fewer than 15 per cent of survey respondents wanted completely self-driving cars.
As well, 96.2 per cent of respondents wanted the ability to take over control of the car, with only 3.8 per cent willing to forgo such ability. Schoettle noted in an email that while the 3.8 per cent falls within the survey’s margin of error, "we do think there would always be a small subset of people who do not care about having controls available."
Such controls are mandated in California, where automated vehicles have been navigating public roadways for some time and where the law requires a licensed driver in the driver’s seat and a foolproof way for the driver to assume command of the vehicle. Is that likely to change as things roll out more?
The report on preferences also asked which interfaces people would prefer. By a slight margin, more survey respondents indicated a touch screen, similar to an iPad or tablet, while a slightly smaller group preferred a voice-activated interface.
For their parts, both Manitoba Public Insurance and Transport Canada are watching developments in driverless cars closely, but have no commitments on policy yet.
"Autonomous vehicles certainly are on our radar, and we’re aware of these vehicles. However, at this point, MPI has not completed the necessary analysis to answer specific questions," MPI spokesman Brian Smiley said in an email. "The reason being, we do not see this as a reality until ‘about’ 2025 in Manitoba. Thus, a little far off at this point."
As for the need for a licensed driver and a fail-safe system, Polzin said it appears the prospect of providing mobility to anyone who needs it will be too attractive to ignore.
"Certainly, the expectation, once it’s mature and fully deployed, is to have people in the cars who wouldn’t necessarily be in a position to assume control," he said.
In other words, soccer moms and dads, you’d only have to go to your child’s practice if you wanted.
Estrada expects parents will, for some time, impose a minimum age for children to take the car on their own. "If the child is, say, 12, and you’ve given him a cellphone and don’t mind dropping him off at the mall, you might be OK with sending him alone in the car.
"I don’t think anyone is going to strap their child into a child seat and send them away, though."
Estrada and Polzin each had encouraging words for fans of motoring fearful of one day sitting in their cars, firmly gripping the steering wheel and doing a full Charlton Heston. ("From my cold, dead hands!")
"There’s a whole culture that has built up around the automobile," Polzin said, referring to racing, collecting and driving cars of all kinds. "That’s not likely to completely change."
John Mahler, chief driving instructor for the Bridgestone Canadian Winter Driving School and chief instructor for Max Performance Driving, isn’t the biggest fan of self-driving cars, but even he admits he wouldn’t mind handing over driving duties during the long, boring stretches of stop-and-go traffic in some major cities.
"I still do worry about how well the software is written," he said, adding self-driving cars will also rewrite the programming for some drivers.
"Self-driving cars will default to the driver’s control if the computer gets confused. If we have a generation of drivers used to being driven by their car, how well will they react when the car doesn’t like its environment?"
Mahler’s question is punctuated by his assertion that driver training and licensing standards are "lamentably low" in Canada, which raises the question of whether the car giving control back to the driver is a good thing.
"It’s not until drivers have had a collision or a close call that they realize how little skill they actually possess. So, automated driving will help many people and until then, the smart drivers will opt for more training."
Mahler said he’d rue the day any government took driving out of the hands of drivers entirely.
"Driving well is a skill that gives great satisfaction," he said. "Any skill that takes training and endless practice to perfect is a joy to the person honing their skill."
The Free Press | Newsletter
Want to get a head start on your day?
Get the day’s breaking stories, weather forecast, and more sent straight to your inbox every weekday morning.
Meanwhile, Estrada, who does not like to drive and won’t miss it ("I’d just rather be doing something else"), said it’s likely we’ll see today’s high-occupancy lanes (HOV) become tomorrow’s automated vehicle-only lanes, with other lanes available for cars with drivers.
He expects in the next three to four years, Google, which is developing its own driverless car, will roll out a version in a small town for testing. As for the day when cars are buzzing to and fro with nobody at the wheel? "I think we can expect to see that in about 10 years."
(Since our interview, Google has taken the first step towards Estrada’s prediction. It has put a fleet of autonomous vehicles on the roads of its Mountain View, Calif., campus.)
He, too, sees the transition as gradual, and likened it to the transition from horses to the horseless carriage.
"Today you go to a dude ranch to ride horses, perhaps in the future, you go to a dude ranch to drive cars."
Mercedes’ sensors offer look into how future driverless cars could operate
The autonomous Mercedes-Benz F 015 Luxury in Motion car.
The beginnings of the technology driving driverless cars are already here, allowing us a sneak peek into the future.
The best example might be from Mercedes-Benz, using the current generation of S-Class, its flagship automobile. The S-Class, as do a few other cars on the road — the Infinti Q50 and Acura MDX to name two — will take over certain facets of driving for you. Travelling down a highway? It will steer for you, for curves up to 30 degrees. Stuck in stop-and-go traffic? It will accelerate, brake and steer for you to keep up with congestion.
Mercedes is careful to point out drivers retain ultimate control and responsibility. Indeed, the system gives you a stern warning if it senses you’re not holding the wheel. But for most situations, that’s for liability reasons, not technological.
Yet the systems Mercedes use give us a look at how the driverless cars of the future will work. The first is GPS navigation, since the car can’t know where to go if it doesn’t know where it is. The second is a host of sensor technologies, from radar to sonar to infrared to a stereo camera mounted near the rear-view mirror.
Radar gives the S-Class its longest look forward: it sends out pulses of microwave signals and reads the reflections coming back. This is good for detecting threats up to 64 metres away. The next is sonar, which sends pulses of ultrasonic energy and listens for the reflections. This is the mid-range sensing system. Infrared looks for heat sources, such as pedestrians or animals. Finally, there’s a stereo camera that builds a three-dimensional picture of the road ahead.
How effective is the combination? It can tell the difference between a human and an animal at the side of the road. If it senses a human on the side of the road at night, it will turn the left headlight toward the person and flash it as a warning. If it senses an animal, it won’t. Mercedes said the system is programmed to know the difference because animals behave unpredictably to a flashing light.
The stereo camera is also central to another technology that might be critical in future driverless cars: Magic Body Control. It’s not available in Canada yet because it’s incompatible with the all-wheel drive system standard in the Canadian-spec S-Class, but it’s really cool and it’s on the road in Europe now.
Magic Body Control uses the stereo camera to read the road ahead and build a computer profile of any bumps or dips that might appear. If it senses a bump, it will, at the precise moment, apply the appropriate force to the wheels (lifting them for a bump, pushing them down for a dip) to very literally make the bump or dip disappear.
Because the expectation is for one day, driverless cars to require zero supervision by occupants, and because motion sickness is most likely when the body experiences forces it cannot reconcile with what the eyes see, bump removal and gentle starts, stops and curves will be critical to preventing AVs (automated vehicles) from becoming motion-sickness machines.
John Estrada, CEO of advocacy group and trade journal Driverless Transportation, said the technology is evolving very rapidly, to the point that if he goes to trade shows four months apart, he’s likely to see a significant advance in technology.
“I was at one trade show recently and Google was showing a car approaching a woman on a wheelchair chasing a duck,” he said. “Now, you don’t often come across women in wheelchairs chasing ducks, but the car did the right thing and stopped and waited for her (and the duck) to cross.”
Most of today’s lane-keeping systems — another precursor to driverless cars — require a lane to be clearly marked for the camera to calculate the car’s position in it. Since driverless cars could be expected to encounter a variety of road conditions, this is one area requiring improvement.
The stereo camera Mercedes uses for Magic Body Control might be part of the answer, as it could provide information about the road ahead independent of line markings.
One of the key technologies still under development is V2V, or vehicle-to-vehicle communication. This allows a lead car, for example, to control the brakes of vehicles in convoy behind. It allows two cars approaching an intersection to anticipate the arrival of each and avoid a collision. The barriers here are steep, but not insurmountable. Security is the key, to make the communication protocols hacker-proof and immune to tampering. Agreeing on a protocol (remember Beta vs. VHS?) will also require either some business-to-business diplomacy or the heavy hand of government regulation.
Fuel will also be an issue: Steve Polzin, director, Mobility Policy Research for the Center for Urban Transportation Research in Florida, recalls a cartoon in which lines of self-driving cars were stranded at a fuel station, since nobody was there to pump the gas.
This could be handled by electric or gas-electric vehicles. Such a driverless vehicle could recharge its battery simply by parking in designated stalls: a coil embedded in the pavement and a corresponding coil on the underside of the car would form a transformer, allowing the car to charge itself wirelessly.