We spent two days at the Detroit Auto Show to experience for ourselves the role that Motor City is playing in the future of mobility. Here are the takeaways:

• • OEMs: Detroit won’t be left behind as traditional auto is consumed by tech. OEMs will continue to acquire or partner with the innovation that will keep them relevant. However, the pace of change would quicken if they would end their obsession with model years and incremental, annual updates and focus on wholesale changes to their business.

• Fleet headwind. As transportation consumption shifts to fleets, consumers will buy fewer cars and fleets will buy more. We believe this will result in a 2% annual unit sales headwind over the next 15 years.
• Suppliers: Internal combustion engine (ICE) part suppliers are in a tough spot as the industry shifts from 1% EV today to 35% by 2030 (Source: Loup Ventures Auto Outlook 2040).
• Dealers: While dealers had a less visible role at the show, it’s clear that autonomous fleets will force dealers to rethink their business model from the ground up.
• Timing: There’s more that goes into autonomy than technology. We are still 3-5 years away from large-scale deployments of autonomous mobility fleets due to the complex web of players that must eventually cooperate. We expect consumer purchases of fully autonomous vehicles to be 5+ years away.

Old school Detroit out in force. While CES presented a view of what’s to come in car-tech, much of the hype in Detroit can best be described as old school Detroit. The show focused on near-term matters like new trucks from Chevrolet, Ram, and Ford, a new hybrid from Honda, and incremental improvements in features. The vast majority of the crowds in the Cobo center were there from automakers, suppliers, and dealers, dressed in suits, there to do business the same way they have since the 80s.

New school Automation growth. The lower level of the show featured the tech-focused Automobili-D, which was full of startups, investors, and universities looking to disrupt the industry responsible for the gathering. Automobili-D, which debuted last year, has already more than doubled in size – a trend we expect to continue as new technology leads the industry’s transition.

Detroit’s industrial complex. President Dwight Eisenhower’s concept of the military-industrial complex (MIC) can be applied to the auto industry. MIC is is an informal alliance between a nation’s military and its suppliers, and those groups work together to influence policy.  MIC-like behavior has been in place in the auto industry since in the 1930’s, which, based on our observations, continues today. So, change will happen slowly with traditional auto, despite recent announcements of heavy investment in the future. Ford announced an expansion of their investment in EVs to $11B from $4.4B and 40 electrified models by 2022, and GM announced plans to build Bolt EV sedans without pedals or steering wheels at scale by the end of next year. These investments represent large bets that should reshape the industry, but the sense at the Detroit Auto Show is that autonomy may be further away than we imagine.

OEMs will play an important part in the future of transportation. It quickly becomes apparent that innovation without the support of major automakers can only go so far. Nearly every startup we talked to was shooting for deals with OEMs or a regulation that will ensure that compatible technology is integrated into all cars coming off assembly lines. Because of their size, history, and longstanding relationships with regulators, OEMs will be the ones to influence legislation or be trusted to perform large-scale tests of new technology. Even the big self-driving players will rely on traditional auto evidenced by Waymo’s deal with Fiat Chrysler and Uber’s relationship with Volvo.

Winners and losers to Detroit’s evolving approach? Traditional OEM’s model will evolve towards fleets, software, and transportation as a service. This means the trend of auto working with tech will continue. We see OEMs evolving away from their legacy supplier relationships and away from the dealer model.  As fleets become more prevalent and individual car ownership declines, so does revenue from new car sales and the subsequent service of those cars. Electric and autonomous vehicles will more resemble a rolling datacenter than a traditional ICE vehicle, so suppliers and dealers will also be under pressure. One opportunity for OEMs lies in transitioning their services from maintaining cars, a hardware business, to owning the future of in-car services like entertainment, a software business. Today, Tesla is the only vertically integrated automaker that is well positioned to take market share away from OEMs given the trends we see in auto. Their absence in Detroit is apparent as they continue to push against the traditional nature of Motor City.

The fleet headwind. As transportation consumption shifts to fleets, OEM’s will sell fewer cars to consumers and sell more cars to fleets. The net of these two trends will be a reduction in vehicles sold, given utilization per vehicle will increase. To illustrate: the average household in the U.S. has 2.0 cars today. If that trends to 1.25 car per household over the next 15 years, that implies about 5m fewer vehicles sold per year in the U.S, resulting in an annual 3% decline in auto sales. Factoring in an increase in sales to fleets should yield a net annual industry vehicle production decline of closer to 2%.

We’re 90% of the way there, but the last 10% is really hard. We continue to stress that our transition to autonomous and electric mobility will take longer, but be more impactful that most of us imagine.

Our transition to autonomous and electric mobility will take longer, but be more impactful that most of us imagine.

It’s easy to imagine a driverless future where there are no accidents, traffic, or stoplights, and everything works seamlessly, but it’s impossible to predict each individual step it will take to get there. Outfitting a car with the proper sensors and building the decision-making software is one piece of the puzzle, but widespread adoption of these technologies will require significant investment and cooperation between countless actors. Young startups, century-old automakers, established hardware and software providers, regulatory bodies from federal to municipal levels, infrastructure contractors, and most of all, drivers, will all need to be on the same page. Today, most of our collective attention is focused on designing cars and their accompanying sensors and software, but other elements like V2X communication, smart city infrastructure, and fleet management must also be considered.

Spending time at Automobili-D revealed the level of effort that is being devoted to each step along the way like interaction with emergency responder vehicles, predictive collision avoidance software, making cars drive more like humans, or simply cleaning off cameras and sensors in inclement weather. We have built cars that can successfully navigate public streets autonomously, but the road to mass adoption is still long. We may be 90% of the way there, but the last 10% may take 5+ years to achieve.