I’m wedged between a BMW driven by a dark-haired woman on my left, and a Corolla with a sign advertising a housecleaning service on my right. Behind me is a man in a Buick whose angry operatic gestures are visible in my rear-view mirror. He will be on his horn in no time. We’ve been creeping along together for twenty minutes, our progress measured in metres. This is Wellington Street, in downtown Toronto. But it’s 2:50 p.m.—not exactly rush hour. And I’m coming from the east—going against the expected flow of outbound traffic. None of us saw this coming, though perhaps we should have.
Toronto, where the average two-way commute is eighty-two minutes, is plagued by some of the worst congestion in North America. In 2016 rankings released by the navigation tech company TomTom, the city took the ninth spot. Vancouver was fourth, right below Los Angeles. Even sunny Ottawa made the list, tying Montreal for fifteenth place.
Few cities around the world are immune from this scourge. In 2010, Beijing had a twelve-day, 100-kilometre traffic jam—and it isn’t close to being China’s most congested city. Even Lhasa, the placid Buddhist capital of Tibet, has vicious traffic.
I live near downtown Toronto and work from home, yet still find myself oppressed by gridlock. It affects where I go, what I do. Whenever possible, I take public transit or walk. My wife cycles eight kilometres to work. Neither of my children has a driver’s licence. As a hockey dad, I am at the mercy of long, punishing drives to suburban arenas. I have sometimes been forced to come to an inexplicable stop at 11 p.m. on the 401, North America’s busiest highway. Where are all these people going?
The average Canadian commuter spends almost 100 hours per year in gridlock. In Toronto alone, the wasted time represents as much as $11 billion in lost productivity and other costs. Traffic-clogged commutes are linked to anxiety, stress, loneliness, obesity, and divorce. (Swedish researchers found that couples with commutes of over forty-five minutes were far more likely to break up.) Car exhaust—a toxic mix of carbon monoxide, nitrogen oxides, and particulate matter—has turned major roadways into so-called cancer corridors, which, according to one study, raise the risk of leukemia for children living nearby.
Traffic has become an existential threat. It makes us poorer, angrier, lonelier, sicker—and it will get worse before it gets better. Perhaps much worse.
Traffic has been with us for a long time. Two thousand years ago, Julius Caesar banned chariots during daytime hours in Rome to ease congestion. In 1879, New York had a five-hour horse-drawn carriage logjam.
Much of the problem stems from our idiosyncrasies as drivers. We switch to the other lane because we think it’s moving faster, even when it isn’t. We rubberneck at accidents, and we’re bad at using turn signals. We accelerate forward optimistically when room opens up ahead of us. We even read signs inconsistently. In one test, drivers faced a sign that read “Falling Rocks.” Half slowed down to look for rocks; the other half sped up to avoid them.
If you’ve ever braked hard in response to an incident—a truck swerving into a lane—you might have triggered a “‘backward-travelling wave” that rippled outward, ultimately forcing cars many miles behind to grind to a halt. In 2008, Japanese scientists performed an experiment in which cars were evenly spaced on an oval track and drivers were instructed to maintain a speed of exactly thirty kilometres per hour. Gradually, more cars were added, one by one. There’s a point when physics kicks in and there are too many cars inside too small a space. But even before that point is reached, drivers become erratic: they slow down, speed up, stop. They bring individual traits—aggression, resentment, poor judgment, marginal skills—to a collective activity.
Being in a car changes who we are. We dehumanize other drivers. We refuse to let them merge; we tailgate and block. We forgive our errors and overestimate our competency: the vast majority of drivers believe they have above-average skills. We prefer large cars or SUVs because they feel safer, though studies indicate that the safer we perceive ourselves to be, the less sensibly we drive. (SUV drivers are more likely to not wear seatbelts and to drive drunk. And sitting up high makes us think we’re going slower than our actual speed.) We are each the centre of our own traffic universe.
The worst manifestation of dysfunctional driving is road rage, a condition some psychiatrists have linked to Intermittent Explosive Disorder (violent responses that are out of proportion to perceived threats). In 2015, a poll found that a third of Canadians admit to suffering from road rage at least once per month. In car-mad China, it has reached epidemic proportions. From 2012 to 2015, Chinese police counted 104 million road-rage incidents. One of the most notorious involved a woman being dragged out of her car and severely beaten by another driver. She suffered fractures and a concussion. When the video of the incident was released online, there was outrage. The male driver, however, released footage from his own dashcam showing the woman cutting him off as she tried to make a missed exit. This triggered a new wave of Internet outrage—this time directed at her .
Making a highway wider isn’t a solution—traffic volume merely expands to fill the available space. When LA tried to relieve traffic on the 405—the most congested stretch of highway in the United States—it opted for a five-year, $1.1 billion project that added a new carpool lane. The result was that rush-hour commutes are now, on average, a full minute longer . Building more roads simply creates more incentive for people to drive.
The truth is, we can’t help ourselves. Our cities were built for cars, and, to a degree, so were we. Italian researchers gave a group of research subjects a number of tokens that they could spend on two travel modes: car or metro. Each mode came with a cost. But while the cost of riding the metro was fixed, the cost of a car changed according to random variables: weather, accidents, road work. Even when the average cost of taking a car was 50 percent more than the alternative, test subjects chose it by a two-to-one ratio.
The bias in favour of cars is more pronounced when transit options are limited. And Canadians’ options generally are awful. Between 1955 and 1977, investment in Canadian infrastructure, including public transit, grew at a robust 4.8 percent annually, keeping pace with population growth and urbanization. But between 1977 and 2001, that figure was only 0.1 percent. The Federation of Canadian Municipalities estimates that the urban transit and transportation infrastructure deficit now stands at $44.5 billion. So it’s hardly a surprise that only 12 percent of Canadian commuters take public transit to work.
There are bright spots, admittedly. Metrolinx, the body that governs transit in the Toronto-Hamilton corridor, has more than 200 infrastructure projects underway, including seventy-five kilometres of light-rail lines. More than $32 billion has been committed. Justin Trudeau’s federal government now plans to double funding to infrastructure over the next two years, with a further commitment of $60 billion over the next decade, a third of it going to public transit.
But even this Herculean effort likely won’t ease congestion. Bruce McCuaig, Metrolinx’s CEO, said the pent-up demand is simply too great. Following the usual pattern observed in other cities, for every Toronto-area driver who converts to public transit, another is expected to take his or her place on the highway.
Political expediency has long been the curse of transportation planning. Plans are made to woo voters and then scrapped by newly elected governments. Toll roads and gas taxes, in particular, are beloved by planners and economists, then dumped by politicians who fear a populist commuter backlash. But things may change as traffic worsens. In LA, a proposed tax to raise $120 billion for transit awaits approval by voters (though Vancouver’s equally gridlock-weary citizens recently voted against a 0.5 percent tax hike linked to transit). The London, England, congestion charge—in effect since 2003—is now $21 per day and has broad support among citizens. Stockholm and Milan have both implemented their own congestion taxes.
“Congestion is never going to go away in the Toronto region,” said Andre Sorensen, a professor of urban geography at the University of Toronto. As density builds in the Greater Toronto Area, traffic pressures will only grow on expressways such as the Don Valley Parkway (built to carry about 60,000 vehicles a day, it currently handles closer to 100,000). Eighty percent of Canadians now live in growing urban areas. No matter how much transit we build, we won’t see much relief in the short term.
But in the long term, seemingly small innovations—such as building more cycling paths—could bring localized relief by encouraging people to leave their cars in their driveways. A year after the installation of dedicated bike lanes (which provide a physical barrier between cars and bikes) in five US cities, cycling volume increased as much as 171 percent. When Toronto implemented separated bike lanes on two east-west arteries, usage more than quadrupled. According to a US Department of Transport study, separating cyclists and motorists is key to increasing bike ridership. Each year, 7,500 Canadian cyclists are seriously injured or killed. The Netherlands, by contrast, has a cycling fatality rate that is one-third the North American rate. The chief reason: infrastructure that divides cars and bikes.
One argument against cycling infrastructure is that Canada is simply too cold for bikes to be part of any congestion-reduction strategy. But Copenhagen, considered the gold standard of cycling cities, is hardly tropical. More than 50 percent of the city’s commuters cycle to work on 400 kilometres of separated bike lanes. (The city is currently building an additional 300 kilometres of “cycle superhighways” connecting the downtown core to the suburbs.) There are other factors involved in this Danish revolution (prohibitive gas and vehicle taxes), but the lesson generally holds that when it comes to bikes, infrastructure matters.
The cycling revolution, should there be one, will likely be led by millennials, who currently make up a third of Canada’s population. For many in that cohort, automobiles are following the same cultural arc as cigarettes: once sexy and cool, now toxic. When they do drive, millennials are more inclined to use car-share services such as Car2go, Zipcar, and AutoShare. (Eletronically connected, they are also more inclined simply to stay at home.)
If even 1 percent of car-driving commuters switched to bicycles, the effect could be profound: a study in Boston predicted that such a shift during peak periods would produce an 18 percent drop in commuting time. And yet, drivers still refuse to ditch their cars. “People want change,” said Eric Miller, director of the University of Toronto’s Transportation Research Institute, “but not badly enough to alter their behaviour yet.”
If our behaviour is slow to change, if politicians continue to fail us, there is always the hope of new technology. Unfortunately, the last game-changing advance in transportation was the Segway scooter, which arrived to great fanfare in 2001 and was reduced to a punchline within months. It was pricey ($9,000), difficult to park, slower than a bicycle or car, not as convenient as walking, and lacked any health benefits. From a transportation perspective, it represented the worst of all possible worlds.
But fifteen years later, a much more promising transportation revolution is afoot: self-driving cars. The modern vision of autonomous vehicles (AVs)—as they are formally known—took shape in 2004, with a first-of-its-kind race sponsored by the US Department of Defense. Fifteen modified vehicles—ranging from Hummers to four-wheel-drive pickups to all-terrain vehicles—squared off in the Mojave Desert. There was only one rule: no human help. Once out of the gate, the cars had to use their own computational smarts to navigate a rock-strewn route filled with steep slopes and other obstacles. Alas, no one finished the 230 kilometre route; no vehicle got any farther than 11.78 kilometres before crashing, stalling or simply becoming navigationally paralyzed. The $1 million prize went unclaimed.
The following year, the prize was doubled. This time, five vehicles successfully completed the race. The top prize was won by a team from Stanford University, whose entry, “Stanley,” finished with a time of six hours and fifty-three minutes. Google hired the team and began developing its own AV. And by 2010, the company had a working prototype.
The US military also got what it wanted: two years ago, it tested its first convoy of unmanned tactical vehicles. More than a dozen firms now have their own models. Volvo is planning a pilot project in Gothenburg involving 100 vehicles. (It’s testing regime is aimed in part at assessing winter-weather performance.) An AV has successfully driven the 5,470 kilometres from San Francisco to New York City, and Google has driven its fleet of AVs more than a million test miles.
Most AVs use a combination of GPS and laser technology called LiDAR (a portmanteau of light and radar). The sensors send out short bursts of light as the car moves, essentially creating snapshots that are fed through the car’s computer and used to construct a 3-D map of the environment. AVs aren’t compromised by variable human reflexes, alcohol, or poor night vision. They won’t slow down to look at accidents, break up with their girlfriend by cellphone while driving, double the speed limit to impress teenage friends, or turn around to yell at kids in the back seat.
The technology is expected to reshape urban life. A 2015 report commissioned by the City of Toronto’s Transportation Services Division concludes that wholesale AV adoption would trigger a 90 percent improvement in fatality and injury rates. The vehicles—which are expected to be commonplace by the late 2020s, according to the report—will ease congestion by moving at optimum speeds and by drawing on continuous traffic reports to determine the best route. Electrically powered AVs would also drastically reduce emissions. If AVs are introduced on a large scale in Toronto, the gains for the city—thanks to reduced gridlock, fewer accidents, and lower insurance—are projected to be about $6 billion annually.
These numbers rely on a key assumption, however: that AVs will reduce the number of vehicles on the road (it is expected that most AVs will be shared rather than privately owned). But Peter Norton, a professor of history in the engineering faculty at the University of Virginia, cautions otherwise. He notes that AVs may merely expand the notion of what we consider to be a viable commute: if we’re prepared to battle through fifty kilometres of ugly traffic while clutching a steering wheel, we could conceivably agree to a faster, safer commute of 100 kilometres inside an AV (especially if we can spend that time working or sleeping). The roads may also be jammed with empty AVs—units that have been dispatched by their owners to pick up passengers or even cargo.
Steven Waslander, an assistant professor of mechanical engineering and mechatronics, heads a lab at the University of Waterloo that builds one-fifth-scale AVs to test the way multiple driverless vehicles interact. I asked him what could happen in the transition phase, when rational robots have to deal with not-always-rational humans on the same roads. He said that when a Google prototype came to a four-way stop sign, it dutifully stopped. Other cars edged forward, a cue for the car to stay put (it is programmed for safety above all else). As a result, the driverless car was continually victimized by more aggressive drivers. So Google developed a car that behaved more like a human; it inched into the intersection, signalling to others that it was claiming its proper place in the queue.
But will that be enough? There’s a lot of instinct involved in driving. Faced with three other cars arriving near-simultaneously at a four-way stop, we humans assess one another. We see what the lasers don’t—that the angry-looking guy with the death metal screeching out of his jacked-up Mustang will likely go first. The retiree in the twelve-year-old Tercel will probably be cautious. We instantly create a hierarchy and act accordingly. There are facial expressions, hand gestures, moments of eye contact—all the semiotics of driving. Computers can’t make these kinds of determinations.
But precisely because most of our information comes through sight, we are compromised by blind spots and the fact that we can see only one thing at a time. The AV can see 360 degrees simultaneously over a greater distance. So that rabid bicycle courier who is weaving in and out of traffic is a surprise to us when he almost takes our mirror off. The AV, however, has been monitoring his progress all along.
In the lab, Assaylbek Dakibay and Arun Das, both twenty-eight-year-old graduate students, talk enthusiastically about the future of AVs. They are currently trying to replace the expensive Google laser sensors with cheaper cameras, thereby bringing down the total price of the vehicles. Arun is also working on a flying AV, a drone that uses cameras to navigate through complex terrain (such as a forest). The model is small, but there are other researchers developing larger passenger models. “I’ve heard of some companies in the Bay Area that are working on flying cars,” he said. “Someone we know is working on it, but he’s not allowed to talk about it.” When I asked them about the possibility of broad adoption in the late 2020s, they felt that was too conservative an estimate: it would happen sooner.
The transportation revolution will be both disruptive and bloody, filled with unlikely partnerships and casualties. Ford is in talks with Google and has plans to create a ride-share program to compete with Uber. Meanwhile, Uber has created UberHop, a multi-passenger service to compete with mass transit, and plans to be driverless by 2030. Both Apple and Blackberry have entered the fray. The AV market likely will resemble the car industry from the early twentieth century, before dozens of manufacturers were winnowed down to the Big Three.
Meanwhile, the existing world of traffic and transit remains underfunded and largely dysfunctional. Until driverless cars can provide salvation, real change will have to come from drivers’ own behavioural shifts, a grudging acceptance of higher gas taxes and tolls, more cycling and walking, and an increase in telecommuting. These solutions are hardly romantic or transformative, but they might be all we have available until the robots take over.
This appeared in the July/August 2016 issue.