Automated vehicles, transit, and physical space

Let’s put the brakes on this one before it gets driven out of control: driverless cars are not the panacea for traffic woes as imagined by some technologists and futurists.

I agree with public transit consultant Jarrett Walker’s assertion that a “completely imagined future” of driverless vehicles is neither realistic nor desirable, given that there will be problems in the intermediary stages, such as how driverless vehicles and human-driven vehicles will appropriately interact with each other, and who will hold legal liability in the event of collisions. However, I am surprised that Walker did not argue the “geometry” angle for why mass adoption of driverless vehicles (particularly Personal Rapid Transit) is untenable – this is one of his significant themes.

In the series on which Walker’s post is based, author Richard Gilbert argues that

Travelling in an AT [autonomous taxicab] will be incomparably better than travelling by bus or even streetcar. As a consequence, ATs will replace much transit as we know it.

No, they won’t. There is the simple issue of the amount of space personal vehicles occupy on the road compared to transit vehicles, whether driven by a human or a computer. Technology, though wonderful, cannot overcome such a physical limitation.

A typical single-occupancy vehicle occupies far more space than a bus, streetcar or train. Even accounting for carpooling (which is not common in North America), public transit vehicles make far more effective use of limited road space than automobiles, reducing congestion and gridlock. While driverless, automated cars are no doubt more spatially efficient than human-driven vehicles and (four times as much, according to Gilbert) and require less space for parking, they still fall well short of the number of people that can be moved in a given physical space using public transit. The alternative, in car-dominated cities, is gridlock, visible and well-documented.

In his defence, Gilbert concedes that heavy rail transit, such as subways and commuter rail, will still be necessary. Transit will actually be necessary on most bus and train lines also, per the geometry issue above. However, driverless cars may be competitive in rural and suburban areas, which are only effectively traveled by car.

More jarring though, are Gilbert’s arguments about comfort and convenience. While “door to door” convenience may be necessary in the suburbs, medium and high-density areas can be very effectively served by public transit, easily reachable by nearly all pedestrians living nearby. There seems to be an underlying connotation of traditional public transit being for everyone else, given that our underlying individualism demands personal space, shielded from the sensory intrusions of strangers. What a sad comment on our post-modern society then.

I’ve used numerous public transit systems – Toronto’s (where Gilbert is based) is reasonably clean, comfortable, and efficient, though the best transit systems in other cities eclipse it. The same cannot always be said about taxis, which driverless cars are primed to compete with.

Putting it on ice

If there is an apodictic truth about humans, it is that we are only human. We have made mistakes, and we continue to make mistakes. Sometimes, we even repeat mistakes that we, or other humans, have made previously.

Arguably, this was the case a few weeks ago in Metro Vancouver, when chunks of ice plunged down upon unsuspecting motorists on the newly opened Port Mann Bridge. On December 19, it was actually cold enough to facilitate some sort of precipitation other than rain. We experienced snow in Vancouver, and the end of the world hadn’t even arrived!

Accumulated snow and ice on the cables of the bridge fell to the bridge deck, some of them smashing into vehicles. ICBC fielded 60 claims related to accidents on the bridge, many resulting from falling ice and snow. One driver was hospitalized.

This was unexpected in winter conditions, even those as comparably mild as Vancouver’s. Except, this type of incident has occurred before. On the same type of bridge.

As the NDP’s transportation critic Harry Bains rightly points out, we should question whether the engineers examined this type of risk on other similar bridges in other regions “like Sweden, Britain, and Boston. Perhaps, also in Seattle.”


As part of a multi-billion dollar highway upgrade called the Big Dig, a ten lane cable-stayed bridge was built in Boston, called the Leonard P. Zakim Bunker Hill Bridge. It is the same number of lanes and basic design as the new Port Mann Bridge.

Since opening in 2003, there have been three incidents of ice falling from the bridge cables on the Zakim Bridge: in 2005, and twice in 2011.

While Boston has a harsher winter climate than Vancouver, we might rightly have asked if such incidents are possible here. After all, it snows in Metro Vancouver, and ice is not uncommon. Despite this, the new Port Mann bridge does not have heated cables to melt accumulating ice, and the engineering firm responsible, TI Corp, has admitted that the coating on the cables did not abate ice buildup. In addition, the cables cross over the bridge deck, which means falling ice and snow is likely to strike the bridge deck, unlike other bridges where the cables connect near the outside edge of the towers and do not cross over the roadway (for instance, the Alex Fraser bridge, also in Metro Vancouver).

So whether the design of the new Port Mann bridge is simply inappropriate for winter climates which include snow and ice (including Vancouver), or whether heated bridge cables would have been an expensive, but practical solution to preventing falling ice, someone made a mistake. And given that it has happened elsewhere (Boston), we should learn from that experience. What is less important than who is at fault, is what we should do to fix the bridge to avoid another ice incident the next time it gets cold in Vancouver.