‘Uber Was Supposed to Help Traffic. It Didn’t. Robotaxis Will Be Even Worse.’
Saturday on San Francisco Chronicle published a joint opinion piece by MIT professor Carlo Ratti (who led MIT Digital Lab research on digital data collection for urban life) and John Rossant (founder of collaborative data sharing platform CoMotion).
They wrote together warning of a future filled with robotics. “Their convenience can tempt us to overuse our cars. The result? An AI-driven traffic nightmare, technically perfect but terrible for our cities.”
Why do we believe this? Because that’s already happened with carpooling. In 2010, the Senseable City Lab at MIT, where one of us is director, was at the forefront of using Big Data to explore how ride sharing can make our streets cleaner and more efficient. The findings seemed astonishing: with minimal delays for passengers, we could find drivers and reducing the size of New York taxi fleets by 40%. More people could get around in fewer cars for less money. We can reduce car ownership and free up curbs and parking lots for new uses. This utopian vision was not only fascinating, but also achievable.
After publishing our results, we began the first collaboration between MIT and Uber to explore a then-new product: Uber Pool (now renamed UberX Share), a service that allows passengers to share cars when heading to similar destinations on a more Low price. Alas, there is no such thing as a free lunch. Our research was technically correct, but we hadn’t taken into account changes in human behavior. Cars are more convenient and comfortable than walking, buses and subways — and that’s why they’re so popular. Make them even cheaper by carpooling and people will be forced to abandon these other forms of transit. This dynamic became clear in the data a few years later: On average, on-call trips generated much more traffic and 69% more carbon dioxide than the trips they displaced. We were proud of our contribution to carpooling, but horrified to see the results of a 2018 survey who found Uber Pool to be so cheap increased total urban travel: For every mile of personal driving it eliminated, it added 2.6 miles of people who would have otherwise chosen another mode of transportation.
With robotaxis on the brink of worldwide spread, we are about to repeat the same mistake, but on a much larger scale… [W]We can’t let a shiny new technology get us into an epic traffic jam of our own making. The best way to make urban mobility affordable, efficient and environmentally friendly is not about new technologies — neither self-driving cars nor electric cars — but old ones. Buses, subways, bikes, and our own two legs are cleaner, cheaper, and more efficient than anything Silicon Valley dreamed of… Autonomous technology could, for example, allow cities to offer more buses, shuttles, and more forms of public transport around the clock. This is because on-demand AV availability can provide “last mile” connections between homes and transit stops. It can also be a godsend for the elderly and disabled. However, any expansion of AVs must be balanced with investments in mass transit and improvements in walkability.
Above all, we need to put in place smart regulatory and tax regimes that allow all sustainable mobility modes — including autonomous services — to scale safely and intelligently. These should include, for example, congestion charges to prevent excessive use of individual vehicles.