Today the California Public Utilities Permission unanimously voted to grant GM’s “Cruise” robotaxi service the ability to charge for rides. The service opened for limited pubic use earlier this year but all rides are currently free. Cruise doesn’t need the money — this is still a pilot project — so they will be charging primarily to figure out how customers react to charging, and what pricing makes sense.
A robotaxi service at first looks similar to Uber
At present, Cruise’s permit only allows operations from 10pm to 6am, and they operate only in the calmer sections of San Francisco, not in the downtown core. Waymo also operates outside the core, but runs all day. Waymo has been charging for rides with no safety driver in Chandler Arizona for quite some time, and received a permit to do the same in SF with safety drivers, but this is the first permit in SF to charge for rides with no safety driver in the vehicle. Chandler, however, is not a robust taxi/Uber marketplace, while SF, the birthplace of Sidecar, Lyft
The following articles have appeared here about this question:
Important questions include just how cheap a robotaxi service can be, if charged by the mile or minute the way taxis/Uber are, and how that gets to being a replacement for car ownership. In particular, it may be that replacing car ownership requires innovative pricing, such as subscription models, or models that mimic the costs of car ownership (which include monthly lease payments, weekly gas fill-ups, annual insurance and maintenance and a few other costs.)
Cruise states that initial fees will be comparable to existing ridehail fees. This only teaches us a little, though one can learn how people contrast a human-driven and robot-driven service at the same price. During development, these services use hugely expensive research vehicles and large operations teams, so they are not actually cheaper to run than a human driven taxi service, but eventually, once all those initial development expenses are covered, a robotaxi can be run for a cost that is based on the per-ride “COGS” plus a profit. That per-ride COGS includes energy and a share of depreciation, insurance, maintenance, cleaning, operations center costs, depots, temporary parking and a few other things.
Eventually this cost should get down to be similar to the cost of operating a private car, which is estimated at around 50 cents/mile. This is less than 25% of the typical cost of hiring an Uber. Compared to Uber, it’s a price that would upend the taxi industry, but might not compete as well with car ownership. Perhaps more interesting is the cost if small, single-person robotaxis are used, if these robotaxis can be built for much less than the cost of sedans, minivans or shuttles like the Cruise Origin. Such rides might sell for 30 cents/mile or less — which is less than even the subsidized cost of a transit ride in most cities.
San Francisco transit fares are $2.50 but the actual cost for each ride has soared to over $14 due the pandemic, from more like $7 before the pandemic. While robotaxis are disruptive to traditional public transit, that is not their primary goal, and indeed they are likely to revolutionize it and make it greener, cheaper and better in every other way — in time.
Cruise plans shared ride service with their Origin vehicle. Today, such services like UberPool (which were shut down during the pandemic) provide rides at a moderate discount to a private ride in exchange for some delays and detours. Cruise has not yet said what sort of service they will offer. Several small companies are trying to operate shuttle service, with fixed stops and even fixed routes, to be cheaper transit.
Taxi services, including robotaxis, avoid the cost of parking and hunting for parking, which can be quite high in a city like San Francisco. In addition, a car replacement service can free up almost 200 square feet of space in a garage in a city where homes sell for $1,200 per square foot! When switching to car replacement saves you $250,000 it can really mean something, though garage space isn’t valued nearly as highly as living space, but some garages can be readily improved.
The real action will come when Cruise, Waymo and others in this game experiment with doing more than copying Uber style pricing. Uber is not their target. The real target is everything about how we get around cities.