Driving revenues: Autonomous cars

-June 10, 2014

In my April blog, I suggested we could see autonomous cars operating in the US before 2020. However, the US may be far from the leading market for such vehicles. In a recent webinar, Lux Research suggested that Japan is likely to lead adoption of autonomous cars through 2030. Further, Lux noted that countries like France are more permissive than the US in terms of providing a path to deploy driverless cars.

Reuters reported last fall that French President Francois Hollande ”turns to robots and driverless cars to revive French industry.” Just a few days ago, Reuters quoted Nissan CEO Carlos Ghosn saying,“The problem isn't technology, it's legislation, and the whole question of responsibility that goes with these cars moving around ... and especially who is responsible once there is no longer anyone inside.” Reuters suggested the statement had to do with Ghosn’s position that Renault Nissan could have autonomous cars ready by 2018, but that could be earlier than legislators and regulators determine how to approve them.

Perhaps Mr. Ghosn feels pressure to keep up with Google, who reportedly is building an autonomous 100-car fleet. As reported by the New York Times, Google’s Sergey Brin decided over a year ago to focus on cars that are truly autonomous, as compared to cars that have a human in the driver’s seat, ready to take over. The issue according to the Times is that the “driver” is likely to be distracted, perhaps reading, which makes them a poor emergency backup system. Thus, Google’s mini fleet will feature cars with no steering wheel or brake/accelerator pedals.

Optimistically, assuming regulations keep up with technology, I was keenly interested in Lux’s estimates of the market value related to autonomous cars, which they peg at a cool $87 billion in market opportunity in 2030. I asked Lux if I could share the forecast details with EDN’s readers, and they graciously approved and provided Figure 1.

Figure 1  Lux Research estimates of market opportunities within the autonomous car market (used with permission)

The first thing that jumps out of this forecast is that software accounts for a huge fraction of the opportunity in the long-run. On the other hand, you might easily overlook the small blue slice that is Lidar (Lidar stands for, variously, Light Detection and Ranging, or Light Imaging, Detection, and Ranging, etc.). Lux said in the webinar that Lidar will be a critical complement to optical/vision systems as well as to radar.

One of Google’s test cars has a roof-mounted Lidar system made by Velodyne Lidar, Inc. of Morgan Hill, California. But Lux said that compared to current generation, $10,000 3D Lidar systems, which include rotating sensors, the future of automotive Lidar will be solid-state devices – without rotation – that may cost $100. Among a handful of companies mentioned in the webinar was small startup Quanergy, a company that says it is a “leading manufacturer of 3D time-of-flight LiDAR sensors designed from the ground up to meet all four key commercialization criteria: Cost, Performance, Reliability, and Size.” I note that in the last several months, Quanergy has filled a few positions listed on their Linkedin company page, including one for a software engineer who by now should be “designing, developing, and maintaining our map data structure and access system and parallelizing localization with a GPU, all based on point cloud data generated by our next-generation LiDAR sensors.” It sounds to me that there will be competition for the segment which is worth nearly three quarters of a billion dollars in five years, and nearly $3 billion longer-term.

Returning to software, Lux said that the “car operating system” could be a differentiator in the future as companies scramble for market share. Of course, such complex software and sensor systems present challenges to validate them and ensure safe operation. Software recalls have been in the news quite often, including a Toyota recall in February affecting 1.9 million Prius models with a software “glitch”, as reported by the Los Angeles Times. I think it is reasonable to assume that software for autonomous operation of cars will be significantly more complex that the various sub-system software components in cars today.

I think it is fair to ask if automotive OEMs can overcome all the hurdles to autonomous cars and see a new era emerge for the automotive Industry. I’m guessing that with $87 billion at stake, they will. Whether or not it proves feasible to test and validate self-driving cars is one of many keys to that future. For example, what happens when a self-driving car encounters many other self-driving vehicles, and through V2V (vehicle to vehicle communications) begins sharing data? Might tomorrow’s cars self-organize into swarms, and if so, would that be good or bad? Fortunately help is on the way. In June the University of Michigan announced the construction of a 35-acre autonomous vehicle test facility with roads, buildings, traffic signals and many other details to present a real-world environment without the risk of testing on public roads.

Also see:

Loading comments...

Write a Comment

To comment please Log In