Google's former Self-Driving Car project, now graduated from Alphabet's X division as Waymo, has found a collaborator and potential new partner in Honda. This is an interesting turn of events given traditional automakers' reluctance to work with driverless-car startups over the years.
In a press release, Honda announced that they're entering discussions with Waymo on how to outfit their cars with self-driving modules developed by Google.
The press release goes on to say that:
This technical collaboration between Honda researchers and Waymo's self-driving technology team would allow both companies to learn about the integration of Waymo's fully self-driving sensors, software and computing platform into Honda vehicles.
As part of the discussion on technical collaboration, Honda could initially provide Waymo with vehicles modified to accommodate Waymo's self-driving technology. These vehicles would join Waymo's existing fleet, which are currently being tested across four U.S. cities.
If both parties agree to enter into a formal agreement, Honda R&D engineers based in Silicon Valley, California and Tochigi, Japan, would work closely with Waymo engineers based in Mountain View, California and Novi, Michigan.
Honda is now the second automaker to announce a relationship with Waymo, following a deal with Fiat Chrysler Automobiles to supply a hundred Pacifica Hybrid minivans, which were debuted a few days ago.
Fiat Chrysler Automobiles (FCA) have been cagey about their deal with Alphabet's Google, with FCA CEO Sergio Marchionne noting that they have yet to determine who would own the data collected in the collaboration.
We need to get to a stage where the car is viable so we can discuss the spoils of that work. We're not there.
None of the other major automakers have partnered with Waymo or Google.
Google's Self-Driving Car project was started in 2009, and in 2012, senior team member Anthony Levandowski (founder of Otto and now leader of Uber's self-driving operations) explained that the team was "talking to every car company to see what their level of excitement is."
For another three years, we heard nothing from either Google or the automakers. From the outside, it looks as if an unspoken cartel was created between the automakers—we'll do this ourselves, no taking a shortcut with outside partners.
Why? Well, the automakers' business model has always been to sell as many cars as you can to each person. The fact that most private cars sit idle some 95% of the time does not bother them—they've grown a successful business model out of convincing everyone that they need their own car.
But when a vehicle drives itself, this paradigm changes, and a single vehicle can now act as a chauffeur to multiple users in the one day. It would then be inevitable that some driverless cars would be used as fleet services, like cabs, and so sales to fleets would increase, but the sales to private buyers would decrease. It's clear to automakers that driverless cars would mean fewer total sales and increased lower margin sales to fleets; a double whammy to the bottom line.
Basically, the automakers have limited interest in the development of driverless technology, and so an unspoken cartel was created. At that time, a fully self-driving system was somewhere between five and twenty years away, and the automakers knew full well that they were still the kings. Why do a deal with an upstart tech company trying to play on your turf? The automakers could carry on making money doing what they do best—selling as many cars as possible—and the upstarts wouldn't have a big brand name or influx of cash or data to grow on.
The status quo remained until December 2015 when a rumor surfaced regarding a potential partnership between Google and Ford. An announcement was expected at CES 2016 in Las Vegas, but in the end, none came. The unspoken cartel remained intact, but the cracks were showing.
If, as mentioned in the announcement, Honda enters into a formal agreement with Waymo, the unspoken cartel may be broken.
A major automaker—Honda—will have the chance to compare their own technology with a closeup of Waymo's technology and find out if their work can even compare. Honda is now on a course to either flirt with Waymo and not commit, or drop their own self-driving technology and go all-in with Waymo. Partnering would get Honda-branded driverless vehicles on the road several years ahead of their rivals (who will need to partner or will be stuck with their long-term project development timelines), and could give both Honda and FCA an unassailable lead.
All the evidence points to Waymo's business strategy as being the mass-production of the self-driving car modules, or the "fully self-driving sensors, software and computing platform" as the Honda announcement puts it. The module is the secret sauce—the decision-making brain of the new driverless paradigm. Honda and FCA now have prime access to the secret sauce, and if it's delicious, expect them to commit to a much fuller partnership deal in the near future.
Therefore, it is now every automaker for themself. Those who have spent huge sums on research and development face a very hard decision—stick with their programs, or consider ditching them for Waymo's module.
The big difference between Honda's and FCA's work with Google is the fact that FCA had not even dipped their toes into autonomous driving research—or any demonstrated autonomous driving capability, as far as we know. FCA directors could look investors in the face and say that they had not spent millions on something that a startup company could do better.
Honda, with one of the world's leading robotic groups, said in October 2015 that they wanted to have a fully self-driving car on the road by 2020, and has been testing their driverless cars at the Gomentum Station facility in California since last June. Honda has invested money into this technology, although nowhere near as much money as automakers like Daimler, Volvo, and Tesla who have longer-running and larger programs. Honda directors can probably justify a possible future partnership with Google without spending too much time justifying their own prior research expenditure in the area.
The Honda collaboration is a watershed moment. An automaker with its own research and development team on self-driving technology has effectively broken ranks and declared the possibility that its own technology might not be good enough or that the project might take a lot longer than expected. To understand the magnitude of this announcement, then an appreciation of the wider automotive and self-driving context is of value.
Some have almost nothing to lose and will be happy to be fast followers of Honda and Fiat Chrysler Automobiles. Expect the unspoken automakers cartel to splinter from this point. We'll likely hear more announcements from major automakers and self-driving startups at CES 2017, so stay tuned.