A city where private cars are largely replaced with driverless taxis doesn't feel terribly close, but it's not too far away to think about. And all the better when someone does the thinking for us—as is the case with this delightfully thorough new report on a "taxibot" world by Luis Martínez of the International Transport Forum (spotted by Gizmodo Australia).
Martínez and collaborators used real trip data to model what life would be like in Lisbon, Portugal, if shared autonomous vehicles of various size ruled the streets. Some scenarios had these taxibots partnering with high-capacity public transit—in Lisbon's case, a subway network—while others considered a transit-less city. (Though taxibots replaced local buses in all cases.) At half a million city and 2.8 million metro area residents, Lisbon is a pretty typical midsized city, though it's car-ownership rate (217 per 1,000 people) is on the lower end of the global scale.
(A quick technical note: We focused on the report scenarios where 100 percent of the car fleet becomes "taxibots," or shared driverless cabs. Other models looked at single-rider car-share fleets, which result in fewer benefits across the board, as well as partial autonomous adoption that mixes with human drivers, which previous work has found problematic from a traffic perspective.)
The good news first—then the bad.
There will be fewer total cars.
Right now about 203,000 cars roam the streets of Lisbon on an average weekday. Even in a taxibot world without public transportation that figure diminishes considerably—to less than 26,000, or about 13 percent the current baseline.
And if the city keeps its rail system up and running, and manages taxibots to facilitate transfers to these high-capacity trains, Lisbon only needs about 21,120 cars. In other words, for every single car in the city that becomes a shared driverless cab, nine other cars can be eliminated.
Here's the best part: Those gains come "while still delivering nearly the same level of mobility as before in terms of travel origins, destinations and length of trip," according to the report.
Rush-hour will get a lot more pleasant.
Taxibots do a pretty good job clearing the road during the morning peak, too. Lisbon's baseline is 60,000 rush-hour cars. If shared autonomous vehicles pair up with high-capacity transit, only about 21,100 cars are necessary for rush hour—or 35 percent of the current fleet. Fewer cars on the road, plus more efficient traffic flows, plus the ability to focus or something other than driving is a recipe for a much more pleasant commute.
There's a lot less wait time, and a little less travel time.
Lisbon's mass transit riders wait about 26 minutes for a train at the moment. In a taxibot world that wait time will plummet to less than 4 minutes—whether or not the rail system survives. Existing drivers will also experience a modest drop in total travel time. Right now the average car trip takes about 18 minutes in Lisbon. Taxibots would shave that to about 15 minutes, assuming transit still exists.
Most of the space needed for parking is set free.
This one's a biggie. Right now Lisbon sets aside space for about 160,000 parking spots. With driverless taxis, which won't need to park anywhere because they're always on the move, the city would theoretically be able to free itself of every single on-street parking spot—space that could now go toward bike lanes or delivery bays or parklets.
Most of the off-street spaces disappear, too, with Lisbon down to just about 8,900 spots (assuming public transit), or less than 6 percent of the current capacity. That would open up the equivalent of 170 football fields-worth of space in the city.
The cars that do exist can get a lot smaller.
In addition to the space saved by the removal of parking spots, cars can become more compact in a taxibot world as well. The report's with-transit scenario finds that about 59 percent of all rides can be served with your standard 3- to 5-person vehicle, with another 6 percent needing a 5- to 8-person large car or van.
But a full 35 percent of all rides can be served by tiny little 1- or 2-person vehicles that only need a sliver of street space. Plus, if they look anything like the "ninja koalas" Google has planned, they'll be friggin' adorable:
Public transit remains critical.
Each of the benefits outlined here gets a little less rosy in a world where driverless taxibots operate without a strong mass transit network to support it. For major cities, eliminating subway or commuter rail or BRT or light rail systems would be a huge mistake. While it seems more conceivable for midsized cities like Lisbon, the gains mentioned above are still pretty persuasive—especially when you add in transit's broader service mission to a city.
And in certain small cities and fringe areas that struggle to support either rail or bus systems, taxibots themselves could become "a new form of low capacity, high quality public transport," in the words of the report.
A lot fewer lives will be lost.
The report itself doesn't address the safety component of a driverless taxi fleet, but given that traffic deaths are one of the leading reasons to pursue autonomous technology, we'd be remiss not to mention that crashes—and crash severity—will plummet.
The one drawback: more mileage.
All the aforementioned gains come with a drawback: taxibots put more miles on the road than exist right now. That's true even when driverless cabs are combined with high-capacity train networks.
Let's take a closer look at the numbers. First rush-hour. Right now Lisbon hosts about 1.04 million vehicle kilometers a day during morning peak. Even with transit connections, taxibots will bump that up 9 percent—to 1.13 million. For the weekday as a whole, taxibots (even with transit) will boost vehicle kilometers from 3.8 million to 4.01 million—a 6-point spike. Vehicle idle time greatly diminishes, too, with the taxibots in motion all but a quarter of the time, compared with the 96 percent spent idle today.
Some of the extra driving comes with the cabs repositioning themselves between trips. A lot also comes from the loss of local buses in this particular report, with those riders now traveling in smaller-capacity vehicles. That could change if buses were part of the equation—though driverless buses themselves are really far off—but considering that other models of driverless taxi systems have reached similar conclusions, it's pretty safe to assume the higher-mileage outcome is a reliable one.
The spike in driving is a net drawback for cities as a whole, but there's hope of containing the impact with cleaner vehicles. If cities were to mandate that taxibot fleets operate with only hybrid or electric cars, for instance, the environmental problems of more mileage would greatly diminish. And despite the additional time needed to charge batteries, the researchers find a "minimal" impact on fleet size (with a rise of just 2 percent).
All the more reason for officials to start planning for the taxibot world well before it arrives.