Autonomous Taxis in Toronto

What driverless Ubers will mean for our city

I remember the first time I tried Uber. My home at the time was a long walk from the nearest transit stop, it was winter, and I was going to a alcohol-infused event. So instead of braving the cold or calling a taxi, I decided to give it a try. This was in 2014, just after Uber launched in Toronto, and almost nobody had even heard of it. The idea of getting into a seemingly random person’s car and having them drive me around seemed completely bizarre and I wasn’t sure what to expect. As the driver pulled up, I remember feeling unsure about the protocol. Should I sit in the front or back seat? Will we have a conversation? Is the payment really going to happen automatically and can I really just get out and walk away?

The driver and I talked about Uber, of course. How cool it was that he had a new, flexible job he could do any time he felt like it, and how excited he was that some people had started to book rides. Neither of us knew that rideshare would become much more dominant than taxis in the coming years. In 2024, there were 212,000 rideshare trips in Toronto completed by over 26,000 Uber or Lyft vehicles. Compare that to less than 7,000 taxis still on the road, down by half in eight years. The numbers tell us what anyone who has taken an Uber knows: rideshare is a far superior experience to a traditional taxi. Easier to book and track, seamless payment of the amount quoted upfront, no need to provide a destination or directions, and a rating system to incentivize good service. The few times I’ve taken a taxi in the past decade have reminded me how much worse that experience is.

While this level of mobility convenience is fantastic and something we now take for granted, another big change is coming. As you may have heard, autonomous taxis are rolling out at scale across the United States, and it’s only a matter of time until they arrive in Canada. Waymo has already begun lobbying the City of Toronto. While there is no way to predict exactly when autonomous taxis will begin service in our city, my guess is probably less than five years, and perhaps less than three. Transport Canada does not yet allow fully autonomous vehicles, but that will change. In the US, Waymo is operating in six cities and completing more than a half-million trips per week with more than 2,500 vehicles. And they’re growing fast, expecting to get to one million rides per week by the end of this year. That’s a lot of autonomous rides, which gives us a pretty good sense of what to expect once those rides begin and scale here.

There are many pros and cons to this new paradigm. The most significant pro is safety. Humans are terrible drivers, evidenced by the 1M people killed globally on the roadways every year, not to mention the many more injuries and even more minor collisions and near-misses. Autonomous vehicles never get distracted or tired, and do not make the kind of mistakes humans do. They are not perfect, but data shows they are much safer. After all, 94% of crashes caused by humans are due to errors such as the driver being drunk, distracted, or speeding. Putting the statistics aside, getting into an autonomous taxi will feel safer because every car will drive the same way rather than the current randomness of what kind of driver you will get.

The next most significant benefit to autonomous taxis is the one thing that did not change when we went from taxis to Ubers: having a stranger in the car. It turns out that mostly everyone greatly prefers not having to interact with, or be in the presence of someone they don’t know while being transported from one place to another. This is one of the reasons so many people choose to drive themselves around rather than take an Uber, let alone public transit. Removing the driver from the equation while still offering the convenience of a taxi is a big deal, as it provides most of the benefits of driving your own vehicle while not actually having to do the driving, deal with parking, refuelling, or maintenance. Riders in San Francisco are choosing to pay 30-40% more than a human-driven Uber just to be in a vehicle with no driver.

There are other benefits, but these are the two big ones, so let’s get to the most significant con: traffic congestion. Just as the shift from taxis to rideshare shifted people from transit to Ubers and Lyfts, even more people are likely to choose an autonomous taxi instead of taking the TTC. Close to half of current rideshare users said they would take transit if rideshare wasn’t available. While not all Torontonians can afford rideshare over transit all the time, there are a significant number of transit riders who have the means to make that shift more often, and likely will. This will not only impact farebox revenue for the TTC, making it more difficult to maintain or expand service, but will also add more cars to our already congested roads. Curbside congestion is also an issue, as autonomous taxis stop for pick-ups and drop-offs, blocking live traffic lanes more of the time. That said, overly congested roads act as a disincentive for people to choose car-based travel of any kind, especially when a viable alternative like transit or cycling exists. The key is to continue to invest in building and maintaining these travel modes, so there are convenient, safe, and reliable options beyond hailing a Waymo.

The city could also tax autonomous taxis to mitigate the congestion issue. The fee per kilometre could be higher when no passenger is present, encouraging the vehicles to operate as efficiently as possible. It’s also possible to mandate or incentivize some degree of integration between autonomous taxis and the public transit system. For example, these taxis could transport transit passengers to their nearest station in areas where density doesn’t justify fixed-route buses.

Winter weather is not something Waymo or its competitors have much experience with yet, as most of the initial cities they’ve deployed in are warm ones. While there will be issues to address, Waymo has the advantage of multiple types of sensors rather than only cameras. Tesla, on the other hand, is camera-only, which does not work very well in a snowstorm. In the long term, our climate will not be a barrier to this service.

Regardless of the potential downsides, autonomous taxis are coming to Toronto, it’s just a question of when. Now is the time to determine the optimal way to adopt this new technology so it complements the transportation system and leads to a better mobility experience for everyone.