Everybody's focused on "sharing" -- is it actually the right first path?

Submitted by brad on Mon, 2020-02-03 09:55
Topic: 
Robocars
Tags: 
forbes

I wrote earlier about Cruise's "Origin" which they say is a vehicle devoted to shared rides. Many other companies also are hoping to make vehicles for shared rides -- it's treated as almost a received wisdom. But the reality is that sharing rides isn't all that it's cracked up to be, and to work what you really need is frictionless instant mode transfers so nobody goes out of their way. And for that you need automated single person pods, not big shared vehicles. In fact, the cost of a human driver, per passenger, isn't that bad on shared riding -- it's in the last mile that the cost is too high.

Read my analysis on Forbes.com at Shared ride vehicles may get it backwards

Share on:
  • Reddit logo
  • Facebook logo
  • Twitter logo
  • LinkedIn logo

Comments

Ken Pyle

Fri, 2020-02-07 09:07

Brad,

Great article and points. I interviewed a Leti/CEATech incubated, French start-up at CES2020 that is pursuing a 1 to 2 person mini-car solution that has many of the attributes you describe. I will shoot you the link when I have edited the video.

Thanks,

Ken

brad

Mon, 2020-02-10 14:26

While I have not written about that particular project, there are many minimobility companies out there, and I have written many other articles on the space.

Bob

Fri, 2020-02-07 17:37

I am surprised you didn't bring up how many AVs might end up driving around empty from one fare to the next. In this likely scenario until deployment is massively scaled, we're going the wrong way. Instead of having only 1 passenger per vehicle like we have today, we end up with less than 1 passenger per vehicle.

FKA

Sat, 2020-02-08 17:40

For this reason alone I don't think it's true that everyone agrees that self-driving taxis are good.

Not as good as long-lasting, self-driving, personally owned vehicles, anyway.

Reducing the number of cars in parking spaces is useful in some places (places where taxis/ubers already are prevalent). In most places parking is not a big deal, though, whereas the number of cars actually on the road driving is (and taxis increase that).

A somewhat hybrid approach is to have short-term leases of vehicles. This allows you to reduce the dead miles, but still allows you to have access to the right size vehicle for your needs. Taxes and regulations might get in the way, though.

Personally I'd probably be fine with just using a four-seater almost all the time, with an occasional lease of something bigger or smaller for long trips. Leasing a one-seater might make sense if I have to travel far away and solo for work, for example.

Once the kids are a little older maybe that'll change. And by the time we have cars that can drive completely by themselves in the places I regularly drive, the kids will probably be much older.

brad

Mon, 2020-02-10 11:24

A lot of study is done by academics on how many vacant vehicles the taxis drive. There have been many results, some as low as 5%, but I think estimates of 10-15% are reasonable. New York Taxis have higher numbers, but they have to cruise the streets looking for fares. Robotaxis sit still waiting for a job, or at most do a bit of predictive movement to get closer to where rides are likely to be.

But if it's 20%, it is a concern, but only a modest one. Particularly if small and electric. And, of course, you can reduce this by increasing the size of the fleet. Larger fleet means a bit more capital cost, more parking cost and longer vehicle lifetimes. You can tune that in contrast with the costs of vacant miles.

FKA

Tue, 2020-02-11 08:53

Even if it's 5% that's more than my current car, which is driving around vacant 0% of the time.

But even 15% seems unlikely outside of tightly packed cities like NYC unless 1) the company's fleet is so large that each driver essentially has a car to himself, or 2) people use the service in addition to having a car (or transit pass) for typical daily driving, not instead of it.

In 2018, ride-hail dead miles were over 34%. http://www.schallerconsult.com/rideservices/favfact1.htm And that's despite most people who use it having other options for their normal daily driving.

Don't forget you have to clean them quite frequently. You have to park them in places where they have permission to park. You have to drive them somewhere to be recharged. These things alone will add significant mileage.

Either way, I don't really see the point. People owning their own cars means a bit more capital cost, more parking cost and longer vehicle lifetimes. And more privacy, more convenience, more certainty, less concerns over the spread of diseases, less overhead, etc. What's wrong with people owning their own cars?

Anyway, my issue was with the comment that "everyone agrees that self-driving taxis are good." No, not everyone agrees with this. There are significant concerns that self-driving taxis are going to significantly increase traffic.

brad

Tue, 2020-02-11 09:24

Of course, while I find it silly, many of the people who expect to own their robocar say they will do things like send it home to drive family around after the commute or loan it to friends, or even the very silly thing of having it drive around the block instead of park, which also would all contribute empty vehicle miles.

The car lasting "longer" is a bug, not a feature. Once a car starts aging with time instead of with miles, that's a cost in your total cost per mile of moving you. And it gets obsolete -- this is more like a computer than a car now, you don't want to expect 20 or 30 years of life for a robocar. Every year it sits underused, it depreciates just by being out of date.

Cleaning is indeed an issue. As for parking, while I think that would be cheap (especially if people use them as taxis and far less is needed) it is also strictly not parking, it is standing, and there are tons of places that standing will be allowed that parking is not.

Not everybody wants to ride in a taxi, but what I meant to say by that line is that in public discussion of the field, you don't usually see people say that taxis are a bad idea.

Jean-Yves

Sat, 2020-02-08 10:51

Your article is spot on and at Toyota we have long thought this was the answer to urban mobility. We did a trial in Grenoble, France with 35 COMS (four-wheeler one pax with a small trunk) and i-ROAD (three-wheeler for two) from 2014 to 2017. COMS are currently operating in Toyota City, near Nagoya in Japan, profitably. Multimodal use is not easy to sell to people but given the right incentives and conditions (key is multimodal payment/pass, I believe) and car bans in city centers, it would work, I think.

brad

Mon, 2020-02-10 11:26

Yes, I have written about the i-Road and vehicles like it a few times. Keep me posted with more as you develop it. I have not had a chance to study the COMS, but without banking wheels its speed is limited, but that can be OK for an urban taxi.

freepress

Sat, 2020-02-08 11:51

What's most striking about this article is that its author criticizes, with zero quantitative rigor, the "room for 4 to 6" rideshare business of a competitor to the author's former employer. Waymo is operating the exact same "room for 4 to 6" rideshare business today and yet gets zero mention in the article. Without a quantitative point, the author is simply trying to hustle sentiment for his legacy and fluff his column spot at Forbes. This article is clickbait.

Why is the "room for 4 to 6" the sweet spot for cities? Because a rigorous analysis of trip data shows it's profitable! DUH! Both Cruise and Waymo enjoy useful *private* trip data from Lyft and Uber. What does such a study look like? Here's a nice one that's based on not trip data but a cell phone proxy: https://www.caee.utexas.edu/prof/kockelman/public_html/TRB18OrlandoDRS.pdf Both the Cruise Origin and Waymo Pacifica are designed to do primarily one thing: maximize the potential value returned to their shareholders. And both companies made those choices through a lot more rigor displayed in this article.

Gosh, can the public replicate these sorts of studies? Yes and no. Uber does offer some public trip data here: https://movement.uber.com/?lang=en-US

It's really really hard to release trip data though because (1) Uber / Lyft don't want competitors tapping into their businesses and (2) properly anonymizing data is very hard, and failure to do so can cause huge problems as exemplified by AOL years ago: https://en.wikipedia.org/wiki/AOL_search_data_leak

While it's nice to have a discussion about the pros and cons of AV rideshare, it's critical to offer something unbiased and with some piece of rigorous substance that hasn't been put out there before. Without authenticity, you're just fluffing your stock options.

brad

Mon, 2020-02-10 14:25

The article puts a focus on Cruise because Cruise made a recent big announcement with a deliberate focus on this plan. Your accusations are inaccurate. Having worked on proto-Waymo 7 years ago hardly gives one a financial interest in Waymo. Owning Google stock does, but Waymo is still only a small portion of the value of that. I make no secret of these things. I am also working on investments that will compete with Waymo. I am bullish on the whole space and feel many approaches -- including shared vehicles -- are worthy of exploration. This article just questions whether the unquestioning drive for shared rides is correct.

Timothy Lee

Mon, 2020-02-10 07:04

"Uber and Lyft invested a lot into UberPool but found that even so, the rides were inconvenient and customers didn’t like them and stopped requesting them. So they raised the prices and people asked for them even less."

What's your source for this? I use Lyft Line fairly regularly and it seems to me that the average number of co-passengers on my trips has been going up, not down.

brad

Mon, 2020-02-10 11:15

It's behind a paywall at https://www.theinformation.com/articles/uber-ceo-khosrowshahi-struggles-to-find-formula-for-success

But that overall improvement masked a worrying trend at Uber’s UberPool service, which lets riders going in the same direction share a car for a lower price than they would pay for the main UberX service. Uber executives originally thought UberPool would one day be bigger than UberX, but instead it has turned out to be a big money loser. To reduce those losses, Khosrowshahi raised its prices considerably—which cut back on the number of people taking UberPool rides. For that reason, Uber’s growth rate for U.S. trips overall has fallen, said people at Uber with knowledge of the situation, although the price increases have accelerated Uber’s revenue growth. Uber rides’ revenue rose 21% in the third quarter but only 11% for the first nine months of the year.

Add new comment