The highway flows smooth and gentle as a river, sleek cars moving in sync; passengers are reading, elegantly sipping coffee, speaking face to face, or tapping away on the next-next generation of phones. Outside, a hazy panorama of vaguely futuristic towers hangs in the background. For good measure, it's also a flawlessly sunny day, without even a hint of a cloud in sight. No more accidents, no more pedestrian fatalities, no more road rage, and no more time lost to rubbernecking and reckless driving. No more traffic jams.   

How will driverless cars change the city? image by UT Flickr contributor Greg's How will driverless cars change the city? image by UT Flickr contributor Greg's Southern Ontario

Alex Mereu smiles as I lay out a vision of a utopian future made possible by driverless cars, hoping to provoke a response from the transportation planner. He laughs. For the IBI Group planner, "it's definitely not that simple," not that I expected it to be. What I didn't expect was a counter-narrative every bit as pernicious as mine was hopeful: induced demand, worsening traffic, declining public transit use, longer commutes, and further urban sprawl. But which will it be? 

The answer is probably somewhere in between. According to Mereu—who recently authored the IBI Group's "A Driverless Future: It's not Just About the Cars" report—the impact of autonomous vehicles on urban spaces will be determined as much by the regulations and policies as the technology itself. "The technology is definitely coming, and much of it already exists," Mereu explains, pointing to innovations like the Google Car and Uber's self-driving pilot project in Pittsburgh as portends of the very near future.

Meanwhile, a number of new technologies have already hit the automotive market in recent years, gradually bringing greater autonomy to vehicles. While some fully driverless cars—more formally known as Autonomous Vehicles (AVs)—now function in controlled, closed-loop environments, a number of more limited features are augmenting the user experience. "There are already cars that drive autonomously on the highway, or that can parallel park," Mereu notes. However, he stresses that the most transformative impacts will come from "Connected Autonomous Vehicles" (CAVs) that are both fully independent—even in mixed traffic—and interconnected with surrounding infrastructure and other CAVs, as well as mobile devices. 

A prediction of the future, image via IBI GroupA prediction of the future, image via IBI Group

Able to operate without drivers in any environment, CAVs would ostensibly be capable of optimizing traffic flow, with the connected vehicles capable of safely and efficiently moving though busy intersections. In this situation, drivers become passengers, able to work, relax, or even catch up on sleep, while their vehicles carry them to their destinations.

For public transit services like the TTC, meanwhile, autonomous technology could help reduce costs in the long-term. Autonomous buses and streetcars could potentially be quicker—though Mereu is quick to note that won't necessarily be the case, especially if they're programmed to be more conservative than human drivers for safety reasons—while also reducing labour costs. (To provide a less exciting related example of automated technology and its benefits, the TTC's introduction of Automatic Train Control allows subways to be run closer together, increasing service). 

For the urban realm, the potential impacts are undoubtedly significant. Perhaps most obviously, cars' ability to continuously operate and serve various passengers throughout the day—before self-parking at a designated facility—could hugely cut down the urban space allocated to parking. Showing up on demand, self-driven fleets of Ubers, Lyfts, and other taxis would carry passengers to their destinations, finding the quickest and most efficient routes, while potentially picking up more passengers along the way, thereby reducing traffic. Notwithstanding moral dilemmas, autonomous vehicles would also probably be safer than human drivers, reducing accidents.

The hazards of driving, image via IBI GroupThe hazards of driving, image via IBI Group

Under a private ownership model, a similarly optimistic vision is being touted by private-sector leaders, including Tesla's Elon Musk. Under the 'Tesla Network' concept, for example, private owners would be incentivized to share their cars—and make money doing so—during periods when vehicles would otherwise sit inactive. Since most urban cars are parked an estimated 95% of the time, keeping them on the road could mean less parking. The popular narrative is that "parking space will be reclaimed for public uses," Mereu tells me, with parks and public markets replacing garages and lots. (In a decidedly less sexy impact, new garages themselves could be built with much lower clearances, since people wouldn't need to stand inside). As the Toronto-based planner is quick to add, however, the outcomes of these car-sharing scenarios are strongly predicated on ownership. How many cars will continue to be privately owned, and how willing will their owners be to share them? 

Even more consequentially, however, making automobile travel easier could encourage more of it. Slate's Henry Grabar puts it more succinctly: "When driving gets easier, people drive more." Just as the advent of the automobile reduced walking, and the demand created by new and expanded highways only served to exacerbate traffic, driver-less cars can worsen existing problems. 

This phenomenon—known by planners as induced demand—can be thought of as a type of collective action problem, wherein choices that make sense on an individual scale become profoundly inefficient when reproduced by the general population. When a new highway is built, offering a more direct route into the city, it makes sense to take it; it's obviously quicker. But when that logic is applied en masse, the volume of traffic makes the new highway much slower, effectively exacerbating the very problem it was meant to remedy.

With self-driving cars, for example, it might make sense for me to take a fast and comfortable self-driven taxi to work, especially if the travel time beats the TTC. Since there's no driver, the lack of added labour cost could make it an affordable ride compared to an existing taxi or Uber. But the paradox of induced demand is that the same incentive structure exists for all my neighbours. It's a great choice if I decide to take it, but at the point that I'm surrounded in the traffic created by my like-minded peers, we might all be better off taking the TTC.

Then, once we do get to our destinations, driver-less, zero occupant cars are still on the road. As impactful as single-occupant vehicles are now, this could pose an even more serious waste of space and energy, depending on how long empty cars remain on the road. 

The potential harms of self driving cars, image via IBI GroupThe potential harms of self driving cars, image via IBI Group

For Mereu, this kind of market-driven model poses obvious risks. If autonomous vehicles serve only the latent demand that exists in the market, the results could be detrimental. "People might also be willing to live further away, and to commute longer, if they don't have to drive," he adds, explaining that self-driving cars could change the incentive structure of commuting, making sprawling communities—and all the negative externalities they pose—more attractive. 

"We also tend to make the assumption that all autonomous vehicles will be electric," Mereu explains, "but that isn't necessarily the case." Although improved technology, changing regulations, and more electric-friendly infrastructure is quickly making these (comparatively) environmentally-friendly cars more appealing to the market, Mereu speculates that autonomous vehicles won't necessarily be electric. Driver-less or not, if gasoline cars continue to be widespread in the coming decades, the environmental impacts will be substantial. 

Ultimately, private cars remain a relatively inefficient way to move people through large cities. Even with no driver and extensive car-sharing, the futuristic taxis envisioned by Uber and its private-sector counterparts can only do so much to mitigate the inhernet inefficiency of the automobile in high-density urban environments. 

Even at their best, shared cars can't do the work of mass transit, image via RPAEven at their best, shared cars can't do the work of mass transit, image via RPA

While private-sector discourse regarding shared, self-driving cars promises a panacea for urban congestion, the reality is that even the most optimistic projection does not prove particularly efficient. Below, an amusing Twitter exchange highlights the limits of the ride-sharing paradise. Self-driven or not, shared vehicles at their most effective and efficient would essentially replicate the services already provide by public transit.

What are the impacts of ride-sharing? image via TwitterWhat are the impacts of ride-sharing? image via Twitter

However, none of this means that autonomous vehicles won't have important benefits, Mereu stresses. "In suburbs and lower-density areas, autonomous vehicles could provide an affordable supplement to limited public transit," he explains. In particular, self-driving cars could pose a reasonably effective solution to the so-called "last-mile problem." In places where transit is relatively accessibly but not within walking distance—such as many suburban environments—autonomous vehicles could provide a more effective solution than the massive parking lots that currently

For people with disabilities and mobility issues, meanwhile, Mereu stresses that autonomous vehicles could provide a huge boon. Efficient ride-hailing and lower costs would increase accessibility, offering a clear and impactful benefit over the status quo. 

***

Ultimately, the ways in which autonomous technology is regulated and guided will be key to determining its impacts. "The best way to predict the future is to create it," Mereu tells me, quoting Peter Ducker. The real challenge, then, is in creating a regulatory structure that accounts for the positive and—in particular—negative externalities that come with private vehicle travel.

As Grabar writes, "[d]riving is too cheap to account for its costs—the deaths, the pollution, the sprawl, the gargantuan investment in roads, and all the wasted time," and the fundamental equation is not erased so much as augmented by driverless technology. For Mereu, this means creating regulations that incentive positive outcomes while discouraging—which is to say taxing—harmful externalities. "None of these are new ideas," he tells me. While the policy proposals depicted below (which include congestion pricing, progressive taxes on low-occupancy vehicles, and incentives for last-mile services) are not without precedent, they could be used to address new contexts.

Potential policy approaches to autonomous vehicles, click for a clear view, imagPotential policy approaches to autonomous vehicles, click for a clear view, image via IBI Group

"We also need a lot of investment in Research & Development," he adds, emphasizing that transit authorities need to be prepared for the coming evolution of travel. Citing Metrolinx's partnership with Rideco in Milton and failed experiments with on-demand bus service in Helsinki as important learning opportunities—"even if it doesn't work out, learning what doesn't work is just as crucial as learning what does"—that can help prepare transit services for the future. 

***

What are your thoughts on autonomous vehicles and the future of transportation? You are encouraged to leave a comment in the space provided on this page, or join the conversation in our associated Forum thread.