What does the arrival of self-driving cars mean for our mobility system? With the rapid development of AI technology, including Tesla's “Full Self Driving”, this question has become increasingly relevant. I recently encountered a presentation by Cern Basher that might serve as a good starting point for this discussion. As this is a topic rich with ongoing debate and emerging insights, I will share my thoughts based on current knowledge. With this context in mind, let’s dive in.
Perspective
Basher is an investment analyst. His analysis seems primarily focused on valuation of Tesla stock, which is fine, he’s doing a thorough job on that. But I think the subject deserves a broader scope than just the angle of a single company, even if that single company is on the cusp of achieving a breakthrough.
My background is that of a transportation and urban planner, so I’m looking at the same subject from a different angle. I think some of Basher’s conclusions are really profound while others look superficial or naive. This is to be expected, as we're looking at an unknown future. Combining different backgrounds and expertise can help. So I will try to find the underlying arguments to explain where I agree and disagree.
A set of technologies exists in a balance with the society that uses them. As technologie progresses, this balance can be disturbed and followed by something else. As an example, in late 19th Century there was a balance around coal, steam engines, steam ships and railways. Mid 20th Century this balance was disturbed and followed by a new balance around oil, internal combustion engines, cars and planes. This completely changed society.
I hypothesize that we are currently experiencing a similar major change in how we use and think about cars, or: a paradigm shift. I'm interested in substantiation as to whether that is the case and what the new paradigm will look like.
The Private Car “Bundle”
The private car was central to the 20th-century technological balance. Over time, as society adapted to both the needs and the possibilities of the car, it has become part of a “bundle” that makes it indispensable in order to participate in society. While we might perceive it as normal, the car industry and adjacent business (like the oil industry) have significant leverage over our lives.
Obviously the emergence of the private car has brought us great advantages. But Basher’s presentation makes clear that it also comes with enormous costs. An average US household spends 16% of their income on transportation. Moreover, the staggering costs associated with inadequate road safety underline these issues.
We can add other notions to that, like air pollution or the livability of cities, and it is clear that the current system is reaching its limits. This is not a new message but it is certainly worth repeating.
Stating the disadvantages of the current system doesn’t mean that it can or should be replaced. A lot can be done to alleviate problems within the boundaries of the current system. In fact, the expectation had always been that we can solve air pollution by requiring “zero emission” vehicles, without any changes to the car system at large. And in my own country, The Netherlands, we have fortunately been able to do a lot about road safety within the boundaries of the system.
The Mechanism of Change
What could lead to such a paradigm shift? Tony Seba's “Disruption” framework explains how major changes occur when new technologies drastically alter the operation of industries. The fundamental aspect Seba examines is the falling cost curve, which he has extensively analyzed for the car industry.
Basher and Seba follow a similar reasoning here, they look at how the per kilometer costs can fall for autonomous cars:
you don’t need to pay for a professional driver, in cases (like ride-hailing) where that currently would be the case;
the utilization of the car can be significantly increased, in cases where it would currently be used privately.
Both work out the business case and show that the outcome is compelling. Current offerings will be undercut on a price basis.
Of course robotaxis will be most similar to ride hailing, which will be affected first. You can discuss about how easily the habit of using your own car will be disputed as well. No doubt many current car users will want to keep their vehicle for practical, privacy or status reasons. It will be a gradual process, starting with those who hadn’t a car in the first place and now don’t find a reason to buy one and followed by those who are happy to gain the cost advantage.
Disruption
In a balance of technologies, different technologies and industries support each other and keep each other alive. But this works both ways, once the balance starts to fall apart, many participants in the system are affected.
Basher gives a clear picture of these knock-on effects, starting with diseconomies of scale for the car industry itself. Whereby the reasoning is actually a bit intertwined between the transition to electric and the transition to autonomous. Due to the availability of new, more attractive cars (or mobility services), demand for the old models will decrease. Thus their production pace will have to decrease as well, which makes them more expensive and brings the industry in a downward spiral.
Basher similarly discusses the effects for auto insurance and service stations. I think the analysis is interesting, the outcome is similar: once the new product takes off, sticking to the old product becomes more difficult and more expensive. Give these underpinnings, the idea that this could lead to a paradigm-shift indeed is likely.
The “Unbundling”
Basher distinguishes between first, second and third order effects. I think this is a wise, but also a bold thing to do.
Given that we indeed see a mechanism of disruption, it is the knock-on effects that will truly change the balance. The third order effects matter the most, they are also the hardest to predict. They change the context from which we try to understand our world, but this is the very same context from which we try to understand the changes themselves.
Looking at the changes from a business perspective might be even more important than the technological perspective. Business chains and “bundles” were part of the old balance, that doesn’t work anymore. Hence the term “unbundling”.
The current “bundle” is defined by car ownership, which forces “one size fits all” for all the trips we make. Since we own the car, it’s this very same car we will use for all the trips we make, be it for shopping, for work or for making a long roadtrip with our family. But once we no longer own the vehicle, the spell is broken. All of a sudden the form-factor that was dictated by the previous paradigm, no longer makes sense. I think Ong and Basher, in the video, are just scraping the surface of a subject that is much deeper.
Yes, of course it is tempting to dream away at the prospect of disrupting other businesses like travel, hotels, real estate and urban planning. From my perspective as a transportation planner, I find these points superficial and naive. As if nobody previously ever has done any research about the connection between transport and society.
Short Distances and Livable Cities
Basher scratches on the surface of trip distances in his analysis of ride-hailing prices. He finds that short trips are both common and expensive, which is reasonably insightful. But he fails to draw the conclusion that the car might be an inadequate form factor for short distances.
Perhaps, at this point, I should mention that I live in Utrecht, a Dutch city that has uniquely prioritized reducing car traffic while promoting active transportation modes such as cycling. Today, bicycles dominate the modal share for short trips. Although we still maintain many parking places and the majority of cars are still ICE vehicles, the implications for the city's livability are significant. The reduction in car trips in favor of bicycles means we require far less asphalt to accommodate traffic, which makes the city greener, quieter, and safer.
Replacing owned cars by robotaxis, without doing anything else, will likely increase the number of car movements because robotaxis need empty trips to go to and from their customers. The Utrecht urban infrastructure will not accommodate those trips, it does not even accommodate current traffic in all circumstances, and rightfully so. If you want to pass through Utrecht by car, you’ll have to accept that it will be slow. The alternative is obvious, take a bicycle, that’s both easier and faster. In the age of “Mobility as a Service” that choice will be even more obvious because you are no longer bound to an owned vehicle. You can take a shared bike instead of a robotaxi.
Long Distances
Then Basher states that long range robotaxis can compete with short-haul aviation, for which he makes compelling arguments. But then, there must be a reason why people use short haul aviation instead of driving their private car right now.
Again, here we are encountering the “one size fits all” restrictions of the private car. It’s not alone that in a private car you have to drive yourself, on longer distances you also start to feel the need for on-board amenities like a toilet or catering. And to use such amenities, you will need the ability to stand up from your place in a manner that is compliant with safety requirements. We need to limit the vehicle’s acceleration. Airplanes (outside of start, landing and turbulence) and trains can do that, for robotaxis it’s not obvious.
Just as with short distances, as an illustration of how far stretching the advent of autonomous driving can be, this is great. But the suggestion that just a standard car, equipped with “Full Self Driving” and a ride hailing contract, will do the job is naive and short-sighted. We may expect an entire new dynamic of offers and competition in this market.
Key Findings
Technological and Societal Shifts: The advent of Tesla's "Full Self Driving" and other autonomous vehicle technologies is potentially leading to a paradigm shift akin to the historical transition from steam to oil-powered vehicles. This shift challenges existing technological equilibria and promises to fundamentally alter our mobility systems.
Economic Impact: Autonomous vehicles, particularly robotaxis, are likely to disrupt existing economic models within the transportation sector by significantly reducing the cost per kilometer of travel. This could undermine traditional car ownership models and favor mobility-as-a-service platforms.
Urban Planning and Livability: The widespread adoption of robotaxis could have profound implications for urban planning and the livability of cities. While they promise to reduce the need for parking spaces and decrease personal car ownership, there is also a potential risk of increasing vehicle movements due to empty trips, thus impacting urban traffic patterns.
Societal Repercussions: The shift towards autonomous vehicles extends beyond transportation, potentially affecting related industries such as auto insurance, service stations, and even sectors like real estate and urban design. Each of these faces both opportunities and challenges as the new mobility paradigm unfolds.
Complexity of Predictions: The long-term effects of autonomous vehicles are complex and multifaceted, involving numerous indirect consequences that are difficult to predict. The discussion on robotaxis is not just about technological innovation but also involves significant economic, social, and environmental dimensions.
Concluding Thought
As we stand on the brink of what may be a transformative era in transportation, it is crucial to approach these changes with a critical eye. While the technological promise of robotaxis and autonomous vehicles is undeniable, the broader implications warrant a nuanced discussion. The potential benefits must be weighed against the challenges to ensure that the evolution of our mobility systems contributes positively to societal goals such as sustainability, accessibility, and urban quality of life. Further research and dialogue are essential to navigate these changes responsibly and effectively.