Big OEMs wouldn't mind a robocar slowdown one bit


We continue to see lots of reports of a "pull back" on robocars, a "winter" of sorts. It is often presented as "the problem was harder than people expected" I discussed some of this in the prelude to my 2018 year in review.

As I indicated, reading these reports it seems like this issues is about 60% due to unrealistic public expectations now facing a reality the industry has already known and mostly been honest about, and the rest a mix of the very common delays in any common software technology, and only a small portion of the "it's harder than we expected" that people are reporting.

The public's expectations were and are due to a number of factors. Elon Musk, with his predictions of "soon, you will summon your car from New York" and general feeling that Tesla is close to having a full self driving car play a big role. Until recently, Tesla even took money from people for "full self drive" to be delivered later as a software update. They no longer offer that on the web site and warn you if you insist on ordering it. As I and many others predicted, it will need hardware upgrades. I think serious hardware upgrades.

There is one factor not being talked about in the idea of a pull-back. Namely that the big car companies would not mind a slowdown at all, and so will happily confirm this narrative, whether it is true or not. The reality is these companies were dragged into this. No industry relishes the arrival of a new technology which will turn it upside down. Those are dangerous waters, and while some car companies will thrive, several will fall. Every sane OEM wants to continue the world where they were making good money selling the same cars they always made to customers, making incremental improvements.

The arrival of electric was forced by Tesla, and the arrival of Robocars by Google. I remember talking to the robocar teams at car OEMs in the early part of this decade. Those teams had existed for some time, but they had modest budgets and were considered far-future R&D. Then Google came out of stealth, and the press were writing "The future of the car is coming from a search engine company." Car OEMs did not like that one bit. So suddenly the small teams became big teams with big budgets. Major efforts took place to sell the message that car OEMs were not irrelevant, that the robocar of the future would come from them, because Silicon Valley doesn't know how to make cars.

(It's true, that Silicon Valley still has much to learn about making cars. But it's a mistake to think it can't learn, or that it can't simply pay car manufacturers to contract manufacture cars. The old knowledge of how to make cars is deep, but it's old and thus for sale at reasonable prices.)

For the car OEMs that have got serious robocar efforts going, they want to be one of the winners, but they would be happy to have more time. Even though people will die in vast numbers manually driving their products while we wait. On the plus side, they are working hard on ADAS to cut down that number in the meantime.

The real "winter"

The small amount of pull-back that is not about public perception comes from a few factors.

  1. We're notoriously bad at estimating complex software timelines. We're bad even when we take into account how bad we are. Even the best project managers look at their realistic timeline for a project, which might suggest it will take 2 to 5 years, and talk about how it can happen in 2 years. It might, but that's not the most likely outcome. In this case, the range is large -- 2 to 10 years, making this worse.
  2. For those with less experience, there is misunderstanding of the 90/10 rule which in robocars is the 99/1 rule. The first 99% takes 1% of the time. The last 1% takes 99% of the time. It's "easy" to produce a car that can drive a nice demo loop and stay in its lanes or drive the highway 99% or even 99.9% of the time. Sweating the details is hard, which is why Waymo is still going slow after 10 million miles of testing.
  3. There are legitimately people who did not understand how hard it really is.
  4. The talk of "Level" 5 as a quasi-real thing blinded some to the reality that these vehicles will only operate in certain roads and driving situations for a long time. It seems that some people thought they were promised this "level" some time in the early 2020s, which was hopefully never promised.
  5. The Uber fatality has raised public fear and with it developer caution. Everybody knew an injury accident would happen some day. An early fatality due to really bad practices was less anticipated, and has, as would be expected, used up a fair bit of the public's tolerance for error.
  6. We're seeing expected blowback from the public as it gets first exposure to real use of these cars. In prototype form -- and in Uber's case hurting people -- you expect the public to react in a way they did not when it was all hypothetical.

The reality remains the same. The best teams will be moving into pilot projects this year and the next, with other teams coming later. Once the pilot projects are underway, we'll learn more about where the remaining issues are and how the public reacts to the technology. While people like to throw out dates, the real release date has always been "when a team can convince its lawyers and board that the technology is safe enough to deploy without creating massive risk for the company." That's a milestone, not a date. There are more date-less milestones, such as:

  1. When a team can deploy safely over a driving region that is commercially viable
  2. When a team is ready to make the massive investment to hit critical mass in that region
  3. When that project proves a workable business model in that region
  4. When the regulations are nailed down
  5. When the project can expand to more than a few cities
  6. When the business model gets accepted to eliminate the need for the early subsidies
  7. When the business model and deployment get good enough to cause significant numbers of people to replace car ownership with ride-based service
  8. When the service can expand to other countries
  9. When the technology can handle new classes of road conditions, such as snow, or chaotic driving

There are more milestones to come. The date of none are certain, though as time goes on we get better ideas about how things are progressing.

And yes, people will continue to imagine fake dates, or believe too much in projections of earliest theoretical dates, and there will be people disappointed.

But it will come. And if the car OEMs slow down too much, they will face the thing they feared when they sped up. Namely that when a big disruption comes, often none of the incumbents survive it.


I'm starting to think autonomous trucking will deploy before Robotaxis. The economics are compelling, even for very limited deployment (e.g. port to distribution center). You can ease into it by starting with platoons. You only need to convince a few customers to deploy and you can prove the business model with a very modest upfront investment.

Robotaxi needs scale. 85 square miles doesn't cut it for a usable service. You need 10k+ cars to cover the Phoenix metro area well enough and with short enough wait times to attract typical customers. That's a $1 billion commitment including back office operations and startup losses. You have to attract and manage tens of thousands of customers instead of just a few. It's a much harder technical problem, with an order of magnitude more variables, for a much riskier business model.

You may be correct, and a few companies have made that bet. I've advised a few of them, including Otto (which turned out to be just a path into being bought by Uber, but I didn't know that.)

The counter is that automating trucking is too economic. You take a higher risk (because the vehicles are big and fast and heavy) and the benefit is strictly making shipping cheaper for the benefit of shippers. Later, there is a good lifesaving benefit since trucks kill 4,000 per year.

But it is some distance from "change how we live" stuff the way taxis are.

It is true as you say that to make taxi work you need a fleet, while a single truck is interesting.

Good points. It's possible truck deployment may speed acceptance in cars, though, and help "change how we live".

There's also the possibility people don't want to "change how they live" by abandoning car ownership, and robotaxis end up mostly pulling people away from mass transit. We won't really know until we try, of course.

And we'll do the trucks too. The thing that always scared me from trucks is the fact that we know there will be accidents. And by the good companies, not just the Uber kind. I want those to be in small cars, not 18 wheel semi-trucks, if they have to be.

And the Teamsters. Employment issues are much stronger on trucks than on taxis. People take truck driving as a career. The don't take taxi driving as a career very often.

No doubt Tesla vehicles will get serious hardware upgrades (especially with regard to processing power) as the costs of serious hardware comes down. But the cost of their cars is already too high for the average consumer, so I don't foresee them making serious hardware upgrades that would require them to significantly increase their prices.

Tesla is making a big bet that there's going to be a big market for people who don't want to give up personal ownership of their primary means of transportation. I think they're right about that. Definitely in the short term. And probably in the long term as well.

Actually, even though they don't sell the $35,000 Tesla yet, even the $44,000 one is not "too high for the average consumer when you factor in"

  • $10,000 in rebates for the average consumer
  • $400 per year or more in energy savings for each year you own it. My electricity bill went down.
  • $600 per year in toll savings if you commute over a toll bridge or paid carpool lane here in the Bay Area
  • Predicted significantly lower maintenance (but more expensive parts in repair at least for now.)

So while not a cheap car, if you imagine 5 years of operation, it's not really that different from many mid-range cars. That's with no $5K autopilot.

Your analysis of the cost seems to be fairly specific to California. I assume the rebates you're talking about are tax rebates. As far as I'm aware I'm only eligible for the federal tax rebate, which has gone down now and I wouldn't have been able to use all of it anyway as it is a *non-refundable* credit. I'm not sure how your electricity bill went down, but presumably that's due to a special program available in your state. And as you point out, the toll savings is specific to your state and in fact only to certain people in your state.

Maintenance will likely be lower compared to an ICE car, but the cost of replacing the battery will be very high.

And yeah, that doesn't include the cost of autopilot, let alone full self driving. Adding an additional $10,000 in sensors would not be viable for a mass market self-driving car, especially not an all-electric one, and especially once the government subsidies run out. So Musk is, I think, going to solve the AI problem of doing it without expensive sensors (which clearly is solvable, the only question is when) or die trying.

But don't take my word on it. Elon Musk said *today*, "While we have made great progress, our products are still too expensive for most people."

Your post's title captures the essence of my impression that the autonomous automobile business is like bicycle racing. People not involved in the sport incorrectly think that guys pedal as hard as they can and the one who can get to the finish line as quickly as possible is the winner. That is completely false in bike racing and AV development. In reality the competitors sit up and look at each other. And look at each other. Until one of them tries to break away. The others chase them down but once the catch is made, they all sit up and resume looking at each other. The sprinters want to keep this laziness up until the last 200m when they spring into action. This is what the major automakers are doing now. They do not want to lose this race, but they do not want to hustle to its conclusion.

While obviously people look at the competition, I am not sure it works as you say. I think Waymo is keeping to its own pace, not worried about what the others are doing. There are things they would have done if they were thinking that way, like start an Uber competitor with human drivers.

What is really happening is that Waymo is 10 miles ahead of the peloton. Everybody else knows they will not catch them by doing the same as them, so they look for strategies that might, if they bet right, pass them. Except there isn't really a peloton. Everybody is scattered on the road. Cruise and Zoox feel they are in 2nd place, but Zoox picked its different strategy before day one. Cruise also picked their different strategy (be part of GM and try to do complex cities early) a long time ago too.

Add new comment