Autonomous driving has become a much talked topic, especially then, when you can dish the dirt in the news. A fatal crash in Arizona? Tesla Autopilot crashes into a fire truck / lane separator / delivery truck? Or just like recently, Waymo cars are still driving clumsily and incalculable, despite millions of driven kilometers.
Fact is that as much as those news – superficially observed – are right. But there is no discussion that autonomous driving will come.
Not everyone understand the same, when they tell us, how far German or other car makers really are. Those ‘magic developments’ that they pursue will ‘soon be here’. But you just need to pay attention to the words used to understand that German talks about automated driving, basically not more than driving assist systems on steroids, while everywhere else we talk about autonomous driving, which does not require a drivers attention or capability to take over control
If we compare the current state of technology and development progress for each manufacturer, we clearly see the stark gaps. While Waymo vehicles on average have to hand over control to a driver every 9,000 kilometers, but in between drive without driver interference in city traffic, GMCruise cars are at 2,000 kilometers until they disengage. But Bosch or Daimler are far behind. Their cars need driver interference every 2 kilometers.
Because a software-based technology tends to improve exponentially, not linearly, those gaps are almost impossible to close. I used the word ‘almost’ just for pure reasons of kindness, the probability that the gap will be close goes towards zero.
Robocar expert Brad Templeton summarized the future of traditional car makers recently in his blog just as a side note. In a more elaborate and less explicit form I wrote about that in my book (Commercial!), but I wanted to dive deeper.
What options for the future do traditional automakers have in a world dominated by robocars?
#1 Become Robocar Manufacturer
Well, that sounds like a self-evident answer, not worth even to be mentioned. But while Mercedes is the pioneer in self-driving, none of the German car manufacturers and only American General Motors as the sole exception has put their focus on self-driving cars. Thee task has long been labeled (and thus dismissed) as too difficult, requiring too much effort, and being too complex. Until the progress that Google/Waymo has shown could not be ignored any longer. While GM has put a large stake into the technology, others were more half-heartedly in their efforts (Daimler, VW), and the rest either doesn’t have money, doesn’t want, or simply can not . This gap is filled by quite a large number of well financed startups, with teams of eager young employees that started putting pressure on traditional manufacturers.
The complexity of the task and the effort to develop self-driving technology cannot be underestimated. It’s mostly a software challenge, and software experienced teams are simply better suited for the task. They code better, are better in data collection and handling, are more willing to test autonomous vehicles on public roads under the eyes of the public, and are more ambitious.
Robocars require a different think on multiple levels. Cars are not sold anymore, they are operated in robotaxi fleets. That changes the sales organization, with the sales model and the competencies required. At the same you car manufacturers operating fleets get into direct competition with their current clients, such as taxi drivers, truck drivers and transport companies. You don’t want to ruin those relationships that were created and carefully maintained over the years. But the biggest obstacle to make that leap is the mindset of employees and a company’s self-concept. More and more people just don;’t want to drive anymore, they want to be driven. Those new startups like Navya have a workforce where up to 80 percent own no driver license! Imagine that. And send that message to employees of a company whose self-image and selection process hires people that have that ‘joy of driving.‘
As you can see, it’s not at all simple to become a robocar manufacturer. And nobody claimed that. But to give up before you start or give it just a lukewarm shot is not helping. You can’t be only a little bit pregnant.
#2 Become Contract Manufacturer for Waymo & Co
Fiat-Chrysler’s and Jaguar’s examples show that the trend of manufacturing cars on contract for fleet operators is already here. Both traditional car makers with a long history neither have the financial resources nor the employees and the knowledge for developing self-driving technology. Instead, they partner with Waymo, building cars according to Waymo specification and equipping the vehicles with Waymo proprietary technology.
For a user it will become less relevant whether they are in a Jaguar or Fiat or a car from any other manufactorer, it will be more important to ‘ride a Waymo’ with the known service experience. For traditional car makers this means two things: brands lose their importance, and their roles turn from OEM to a supplier.
What may sound like a bad deal for today’s OEMs may turn out to be the only chance to move into the future and exist.
#3 Vendor for Niche, Which Is Not Yet Dominated By Robocars
Initially, robocars will occupy certain segments. In urban areas for instance, where public transport, taxis and Ubers today are providing transport solution, robocars will increase their importancee and reduce demand for privately owned cars. Rural areas may remain for the immediate future the space where privately owned cars can still be sold to individuals. Or to people that do a lot of driving for their jobs.
But sooner or later also those areas will see an increase in robocar use, as soon as their reliability and the capacities are here. One reason will be the price. Many industry experts expect that the price per mile will decrease by at least 50 percent to what we pay today. And when heavy drivers who put a lot of miles on the odometer for their job will be able to relax or prepare the meeting with their customers while they are on the road, then even those niches will be lost forever to robocars.
Those nice markets will be so small in less than 15 years that no car maker without self-driving technology will be able to operate profitably. Similar to coach makers, Polaroid cameras, or vinyl record companies 90 percent of those companies will disappear. Those companies suffer from fate #5.
#4 Vendor for Niche, Which Will Never Be Dominated By Robocars
The remaining 10 percent of companies, who survived #3, will operate in that spacee. Some vehicles will never see a robot at the controls, and some will be used for pure leisure. Driving such vehicles manually on race tracks on the weekend or a closed for public road along the coast or up the mountains will be the niche. Similarly to coaches that you use today for weddings, competitions, or as fun for tourists in a city center, but that you never see on highways, interstates, or cross country roads.
#5 Demise / Death
Fifty to 90 percent of companies will suffer from that fate. In the best case they will be bought by robotaxi manufacturers, in the worst casee they have to shut down. Or they exit car production completely and do something else.
Do 50 to 90 percent seem way overboard? Well, then consider the following: what cell phone brand did you have 10 years ago? What brands do you remember? Nokia, Samsung, Ericsson, Siemens, Blackberry, Motorola. How many of those companies have a significant market share in smartphones today? Just Samsung survived that space, the rest is history.
Whom Will It Hit?
What manufacturers will be suffering from what fate? Of course predictions are difficult, especially when we talk about the future. But we can see signals, if we just bother looking, and from those signals we can build scenarios for each car company.
The irony here is that Daimler is a self-driving pioneer. In the 1980s they did their first astonishingly ambitious and successful test drives. From Stockholm to Paris through all of Germany went their experimental drives, with trunks, back seats and trailers packed full with hardware and cameras. But it was still too early for the technology available, but still, that was quite a feat.
Thirty years later it’s almost embarrassing to realize how far Daimler has fallen behind. While Waymo disengagements are on average every 9,000 kilometers, Daimler’s vehicles have to hand over control every 2 kilometers. If you understood that this technology develops exponentially, not linearly, then you realize the meaning of the figures. As soon as a company has that head start, they become almost out of reach for everyone else.
Daimler is a company that – in comparison with other German car manufacturers – invested the most in new technologies. And still there is little to show for. They have done enough to fall into the belief that they understood the technology, but accomplished too little to become a serious competitor for the likes of Waymo.
But that’s not completely understood at Daimler. The belief is that partnerships with B-Players are enough (for instance with Bosch), while Daimler considers itself as an A-Player and thus should be partnering with other A-Players, like Waymo. But they seem too proud.
I’d give Daimler the biggest chances to survive in an orderly matter and without fire sales. But not as an independent car manufacturer, but as a Tier 1 supplier for the Waymos, Apples & Co, or as a subdivision of one of the latter. The reason for my optimism is the aspiration to quality and corporate culture that would fit with buyers such as Apple. The probability that Daimler turns into a robotaxi fleet operator are to my estimate below 50 percent. But even then Meercedes will be to Waymo – in regards to market share and significance – what Bing is today to Google.
Where should I even start? BMW has not much to show for in regards to self-driving cars. The BMW CoPilot is a souped-up driver assist system. If they are lucky, they’ll be acquired by Waymo, Apple & Co. With less luck they become a Tier 1 supplier, and with a lot of bad luck they are being acquired by a competitor aiming at staying in #3, and thus get wiped out together with them in #5.
The technology lead – as they call it at VW – for autonomous driving so far was Audi. And they showcased a very fast A8 driving on a race track. But that’s also where the excitement stops. Instead, VW entered a partnership with Aurora, a startup founded only a few months ago by former Waymo chief engineer Chris Urmson. And the Aurora-founders certainly have the competency and drive to build that technology. It speaks for Aurora that they refused an acquisition offer from VW from the get-go.
But Volkswagen’s plans seem rather uninspiring. The first tests with robotaxi fleets in up to five cities are not expected to begin before 2021. Up to this point Waymo alone will have tens of thousands self-driving cars in fleets on the roads. As mentioned multiple times by now: this development is not linear, it works exponentially.
With all the troubles that Volkswagen heaped on itself (Diesel scandal among others), and the still ongoing reverberations throughout the company – now with the new CEO Diess also coming under Scrutiny and the Audi-CEO still in custody – more fines are unavoidable. The focus is anywhere else but at self-driving cars. That will be punishing them later.
The scenario will be one, where a money shortage will force them to sell units such as Porsche, MAN, or Bentley, in order to raise the money to build cheap, mass ready robotaxis on a base of the Beetle, and a Golf/Passat. They are likely to license the technology from the likes of Waymo and pay them billions in license fees every year.
I purposely took Porsche out of the VW-group, as this is a bit of a different car manufacturer. You buy a Porsche less on a rational base, but because of the pleasure of driving it.
While Porsche has its promising first electric vehicle named ‘Taycan’ (formerly known as Mission E) right on schedule for next year, the company faces the dilemma how to convince customers of the usefulness of a self-driving sports car. The fun of owning a Porsche is in driving it manually. But the signs are on the wall. Around 2030 manual drive will be outlawed on public roads. Due to safety reasons it will be illegal to manually drive in public, only self-driving vehicles will be permitted.
For Porsche to survive as a niche vendor as sports car maker, they are simply too large. But Porsche has already some time ago left the sports car focus. With the SUVs that the company sells they still bath in that glory and sell them well, but those cars a comfortable city dwellers with high workmanship quality.
Porsche today is dependent from other efforts inside the VW-group, waiting for Audi technology (that screwed them with the Diesel in the past) or the Aurora partnership. But all that doesn’t stop them from doing their own technology developments.
Porsche;s survival rate might be higher than 50 percent, but not inside VW, but as a luxury manufacturer for the Apple & Cos.
The fate of Opel is easy to draw: #5
Here is a summary of the fate comparisons with an informed estimate from today:
- #1 Become Robocar Manufacturer
- #2 Become Contract Manufacturer for Waymo & Co
- #3 Vendor for Niche, Which Is Not Yet Dominated By Robocars
- #4 Vendor for Niche, Which Will Never Be Dominated By Robocars
- #5 Demise / Death
Licensed from Waymo & Co
What do you think? Post your opinions below in the comment section.
This article has also been published in German.