The California PUC has issued a permit to Cruise (funded by GM and Honda) to offer robotaxi service in California with no safety driver in the vehicle. Cruise tests in San Francisco, and touts the importance of developing in that more difficult non-suburban environment. Waymo has yet to request a no-fee permit, but reports suggest that both Waymo and Cruise have started the process to offer for-pay rides. Waymo operates a for-pay, no-safety-driver service in Chandler Arizona and has done so for some time. It did not bother to request a no-fee permit for California, but has been ramping up operations there, suggesting it hopes to do that soon.
It’s not that surprising that Waymo has not bothered with a no-fee permit, considering the limitations in the Cruise permit, which also forbids return trips, airport trips or shared rides. Cruise, on the other hand, is yet to do either public taxi service (with safety driver) and only did one brief demo of a test ride with a safety driver in the passenger seat, late at night in a quiet part of town.
Testing with passengers is an important step, and you learn about things like:
- How people react to vehicles and what they do in vehicles.
- Ride-ordering, management and (eventually) billing user experience.
- How to provide customer service from remote operators.
- Dealing with and learning about strange customer requests, like changing their minds, breaking rules and many other things.
- The challenge of pick-up and drop off, which is also quite different in suburban strip malls and urban curbsides.
- If you can charge, how people react to fees, and what fee structures work best.
Some of these things companies like Uber
Cruise has said their real plan is their shared “Origin” vehicle but they can’t test ride sharing. San Francisco is finally getting ready to be enough beyond Covid to bring about the return of ride-sharing, and it’s a shame to slow down more experiments in that area, too.
Waymo has so far only done modest fare experimentation. Their fares in Chandler are a modest discount to Uber fares. It’s not clear that this affects how much people use the service there. Uber-style riding is vastly more common in San Francisco than Chandler. Waymo recently added the ability to do a multi-leg ride (including of course, return trips) by pausing your ride, with no need to pay the “flag drop” fee for the extra legs, just adding on more miles.
Typical taxi rates will include a base “flag drop” fee (a reference to the flags on taxi meters) plus a per-mile fee and sometimes an idle-minute fee for being stopped. There can also be a minimum fee, which in the case of Uber is often around $7, making short rides uneconomical. The flag-drop pays for various things, including the times drivers sat around or cruised waiting for passengers, and the distance driven to pick up a passenger. Robotaxis allow experimentation with this. There is no human who needs to be paid to “wait around,” just the lower cost of keeping a vehicle idle. The costs will largely be per-mile, but something has to pay for the trip to pick up a customer from the last drop-off, and any predictive repositioning to make pickups faster.
A lot of the interesting experimentation to be done with robotaxis goes far beyond this. Customers may prefer flat fees or monthly subscription fees to per-mile costs. Uber and Lyft have toyed with subscription services. Regular car ownership isn’t seen as costing per mile — for many owners, it involves a monthly payment for lease and insurance, maintenance payments 1-2 times a year and weekly visits to a station for tank of gas (or monthly electric bills.) Plus parking and garaging of course, something taxi users don’t worry about. The taxi style of payment is foreign to the car owner.
Waymo still has more to learn about driving in San Francisco, as does Cruise. They both have to learn about crowded curbsides at busy times. But a lot of the interesting learnings involve payment. It is not at all clear why the California PUC wants to forbid charging, and in what way they are fulfilling their role of protecting the public by doing so. These services are both going to start small, so they will not (as yet) affect the public markets for rides, though their long term goal is to outcompete them. Let’s hope they are quick at reversing this policy.