When self-driving cars arrive on the streets, one early beneficiary is likely to be congested and polluted cities. Shared, automated electric vehicle fleets promise to help city leaders reduce traffic, improve the environment and deliver reliable, on-demand transport to residents.
To investigate this shift further, a team of researchers from the Department of Energy’s Lawrence Berkeley National Laboratory and the University of California, Berkeley (UC Berkeley) recently analysed the cost, energy and environmental implications of a fleet of self-driving electric vehicles operating in Manhattan.
In our latest interview, we find out more about their work from Gordon Bauer, a Ph.D. student in the Energy and Resources Group at UC Berkeley and one of the authors of their study – “Cost, Energy and Environmental Impact of Automated Electric Taxi Fleets in Manhattan”.
What led you to look at the implications of a fleet of self-driving electric vehicles operating in Manhattan?
“For several reasons, self-driving fleets are going to be very important for implementing electric vehicles on the scale needed to prevent catastrophic climate change,” says Bauer.
His colleague Jeffery Greenblatt had previously demonstrated in a paper for Nature how the economics for electric vehicles become more favourable the more they are used, thanks to lower operating costs.
The team therefore decided to focus on Manhattan for its case study, “because taxi services in Manhattan are already extremely common and owning a car is very expensive,” says Bauer. The area’s density means that, once the technology is ready, it could be one of the first places to see shared, automated electric vehicles on a large scale.
To help construct a realistic model, the researchers harnessed detailed taxi trip data from New York City’s Taxi & Limousine Commission.
“It could be one of the first places to see SAEVs on a large scale.”
Can you tell us a bit about what you found?
Self-driving taxis promise to radically reduce the cost of taxi services, says Bauer, “which could radically change the way people get around in urban cores.” At that point, he questions, “what happens when taking taxis is cheaper than owning a car, and close in cost to taking public transit?”
Electric vehicles could also result in significant cost savings, found the team. Their model differs from some other studies, due to the flexibility it allowed in battery range and charging timing. But if taxis can charge whenever they are idle, says Bauer, “we found that the lowest-cost fleet would have around 50-100 miles of battery range, without any fast-charging infrastructure”. Of course, he notes, this assumes taxis would have access to automated charging stations throughout the city.
The team also found that ‘range anxiety’ is unlikely to be an issue for shared city fleets. While it matters to those looking to buy a car or who are worried about finding charging stations, “the average taxi trip in Manhattan is about two miles long”, says Bauer. “Except for a short time during rush hour each day, some vehicles will be free to go charge.”
Did any of the findings surprise you?
Bauer was very surprised by which components of cost were most important. “Vehicle purchase is a large part of it, but it doesn't change much when you vary fleet parameters,” he notes. When there are more vehicles in the fleet, each one lasts longer and the minimum fleet size is constrained by the number of vehicles required to serve peak demand.
“Meanwhile, everyone thinks charging infrastructure will be really expensive, but when you compare it to the cost of buying, operating, and parking thousands of vehicles, it's actually pretty small,” he says.
“I look forward to the day when car crashes are as rare and as tragic as plane crashes.”
What infrastructure challenges could the need for a fairly dense charging network present?
From an engineering and cost standpoint, says Bauer, it would be best to site chargers in a dispersed network throughout the city. “This could be done at a relatively low cost,” he notes, but would present major political barriers such as deciding who would get access to which charging spots, and removing public parking spots in downtown areas.
The alternative model is to charge vehicles with super-fast chargers situated in large warehouses on the outskirts of cities, explains Bauer. “But this increases cost, and also could lead to significant empty vehicle miles driving to and from charging.”
Either way, he says, “electrifying transportation will require significant changes to the power grid”.
Is there anything city leaders should start doing now as they plan for SAEV fleets?
The main barriers to effective implementation will be parking permits and charging stations, says Bauer, especially in areas with high demand, such as city centres.
In particular, he says, “if cities are going to effectively manage congestion and incentivize ride-pooling, they will need to implement per-vehicle congestion charges.”
Which part of SAEVs – the shared, automated or electric element – will be most important when cities are considering such fleets?
Bauer says that when it comes to the element that will have the greatest impact on society, it’s “definitely automation”. However, in terms of which one is most important in realising the potential benefits, “I’m not sure you can have one without the others,” he says.
“Unless vehicles are shared, congestion could be a nightmare: imagine having your car drop you off at work, then driving back home to avoid parking fees. Unless they are electric, the environmental impact could be a disaster. And unless they are automated, sharing vehicles will be pretty expensive, and charging EVs will be inconvenient.”
Ultimately, says Bauer, “without self-driving, taxis will never become a dominant mode of transportation, and charging will likely remain a major barrier to EV adoption.”
Your study used Manhattan as the model – are there other cities that this model would also apply to?
The team’s model “certainly” applies to other cities, says Bauer. While Manhattan’s density provides a certain advantage, “there is nothing unique about it”, he says. “We have also modeled automated electric taxi fleets in multiple cities in China, and found similar results.”
The team is in the process of expanding the model to other US cities, and is looking at scenarios where demand for shared, automated electric vehicles “goes beyond present-day taxi demand to include a high percentage of all driving trips”.
Finally, what are you personally looking forward to most about the advent of self-driving cars?
“I look forward to the day when car crashes are as rare and as tragic as plane crashes,” says Bauer.
About the expert
Gordon Bauer, Energy and Resources Group, UC Berkeley
Gordon Bauer is a Ph.D. student in the Energy and Resources Group at UC Berkeley. His research focuses on leveraging shared, automated, and electric vehicles to reduce social inequality and environmental impacts, drawing on insights from a variety of disciplines to understand how people might interact with these new technologies.
We speak to Prof. Dr. Andreas Herrmann from the University of St. Gallen about his new book Autonomous Driving: How the Driverless Revolution will Change the World.