Google gets a bunch of news -- and that is news. Also, contract manufacturing


There have been recently a few news announcements and in depth coverage of Waymo.

These pieces are interesting if you follow Waymo, but they, along with a number of other recent stories, signal a major change at Waymo -- they are actively engaging the press.

Waymo has never totally hidden from the press, but as part of Google/Alphabet, they have never had the need for press that other ventures do. Google is the world's #1 brand. Anything it does gets tons of press automatically. They don't talk to the press unless they want to.

So this means they now want to. Which means they have made a real decision to take it to the next level. Of course statements like "million rides a day" say that explicitly, but lots of companies have made bold predictions of what they plan to do and why. Waymo has declared it is ready for public engagement and scrutiny, and wants to mold the story. In effect, even though they were not in stealth, it is like they have come out of stealth.

Car makers waking up

On another note -- probably related in an obscure way -- I heard from a source I won't name that one of the major automakers has decided it will be quite happy to be a contract manufacturer in the robocar world of the future.

Of course, FCA is making minivans for Waymo, and Jaguar has an order for 20,000 E-Pace vehicles. Volvo has an "order" for 24,000 Volvos as well. But there has always been unease at the idea. Dieter Zetsche of Daimler has declared for years that he doesn't want to be the Foxconn that makes the cars but doesn't design them, though I have predicted that will be the fate of several car companies.

Indeed, there are only 5 fates I see:

  1. A leading robocar company
  2. The contract manufacturer for a fabless robocar company
  3. A non-robocar maker doing well in markets not yet dominated by robocars, leading to #5
  4. A minor player doing non-robocars for niche markets
  5. Dead

Since only a few players will survive to slot #1, it makes sense that players should get ready to accept choice #2. It's not as high margin, but it's not a terrible place to be. If you focus on being the best at it, you will do very well, because there are going to be more robocars made than there are regular cars made today.


""because there are going to be more robocars made than there are regular cars made today."" I agree with your 5 points.
If we try to make an agreement between all parts interested, in no more than 10 years we can have only autonomous vehicles (car, motorbike, truck,bus), worldwide.
ZERO human driving vehicles, with near ZERO deaths because accidents,which is the most important.

That does not happen quite that fast. For one thing, people who don't live in cities -- that's a lot of people -- will not get on demand taxi service, and will continue with car ownership, though in time it will be robocar ownership. Not in 10 years though.

1)The main problem about a fast substitution of HDV (Human Driving Vehicle) by RC, is, if it is possible in 10 years to produce RC to match the same amount of HMV we have today.
You have the proper info, but roughly, every year is produced 1/10 of the circulating vehicles world wide (do not count motorcycles).
2) The substitution should be done area by area. Totally in each area, in a few days. This will eliminate the problems of coexistence between HDV and RC.
3) The problematic areas for RC, will be the last ones, which will allow them to be adapted to RC. (like rural areas)
4) Some type of organization will be responsible for the financing, maintenance, administration of trips, etc. of the RC fleet, which will avoid accidents and circulation problems due to bad attention of the RC by its owner.
5) Each user will pay the use of the RC for time, distance and other criteria.
6) You can be "owner" by contracting a RC for a period (day, month, forever).
7) All the inhabitants of the area, permanent or visitors, will be able to use the RC service, independently if at any time or always they do not have money.

8) These are basic starting points. Each one should be expanded and solved the questions.

Do you think all the main technical challenges have been solved, and "un-stealthing" means the next steps are business/infrastructure/marketing buildout? If not, what technical challenges remain?

Waymo, it seems, has solved the problem of driving in Phoenix. However, there are other areas with lots of special challenges and complexities -- Phoenix is about as easy as it gets. Driving in snow is one example, which they have said they are working on.

First you have to solve it, then prove to yourself that you have solved it -- the latter is often harder.

Non-robocars will continue in some form - Lamborghini, Ferrari, Porsche, Honda, Hyundai, Subaru, Land Rover, Jaguar. No robotaxis for the sporty marques!

But don't be surprised if your Porsche is able to drive you through boring city traffic to the hills outside of town to let you take the wheel on the fun road.

There is a truth hidden in Brad's prediction. The current cars, mainly special models, Ferrari, Porsche, etc., can be produced, if they have demand, but with an autonomous control system, which will allow manual operation, only in certain areas, enabled for this purpose.
A Ferrari physically equal to the current, even its interior, with identical engine, could be acquired for ever or rented for a period.
But it will also have a system of autonomous management (possibly with electric motor).
Within the RC areas, the Ferrari will be RC. The RC system will inhibit that the Ferrari can be driven by a human in those areas.
No one will be able to drive a car manually in an RC area.
Only RC in an area, does not prevent you from circulating or possessing the type of car you want, but in the area it should only work as RC.
Today you can have a horse and maybe elephant or camel in your yard ....but not drive it in many areas.

I predict that in a few dozen years, driving a car will be quite a bit like riding a horse is today. Plenty, but not most, people will do it, and it will mostly be a moderately expensive hobby.

Each city a Waymo comes into will mean a surplus on the used car market. People will get rid of their cars in favour of Av services and demand for brand new cars will dry up if the used car market has excess supply.

Once the self driving software problem is solved, many other software companies will solve it as well. This means enterprising people will be able to outfit these used cars with cameras and a radar, licence the "Linux/Androidish" version of AV software, and get their own robotaxi outfit going.

This will be bad for hardware manufacturers because we will reduce the need for new steel frames. Let's see how each country regulates the shape of AVs, but my bet will be that we get rid of the car shaped steel frame eventually and have something more like a moving room on wheels. Their factories are not tooled for vehicles like that so they will be ripe for disruption.

People will sell some cars, but they just sell them to cities, and later countries, where there is no robocar service yet, and to the people who want a good deal on a manual car.

The problem with the open source version is the safety certification the government will demand. I have put Linux on dozens of computers in my day, random configurations of parts. If there was some sort of complex certification needed before I could run it, I would never have done it. Just bought a new certified computer.

Add new comment