In January 2020, Tony Seba gave the keynote speech at the North Carolina Department of Transportation 2020 Summit.
He titled his speech "The Future of Transportation."
It is a very good speech. If you have an hour to watch it, I recommend doing so. Or, you can just read my summary and get the key points he made. This summary is not a transcript. I have quoted some of what he said, but mostly I've organized and paraphrased his words.
Tony begins talking in general about technology disruption. He begins with the adoption of automobiles. He showed a picture of the 1900 Easter parade in New York City. The road was filled with horse-drawn carriages and one automobile. He then showed a picture taken in 1913 in New York City. The road was filled with automobiles and one horse-drawn carriage.
He gives the following definition:
"A disruption is essentially when there is a convergence of technologies that make a product or service possible, and that product, in turn, can help create new markets ... and at the same time destroy or radically transform existing industries."
Tony explains a few case studies of companies that have experienced disruptions. These disruptions were often glossed over by analysts. He cites AT&T and Nokia.
He asks, why do smart organizations fail to anticipate or lead in technology disruptions? He talks a little about his disruption framework, which he has created to examine disruptions. He asks, "can we anticipate, can we forecast, more or less, disruptions to come?"
Tony takes a dive into technology cost curves. As production expands, the costs of technologies come down. Technologies are not adopted linearly but are always adopted in an S curve manner.
He reflects on the speed of the automobile adoption. It went from 0% to 95% nationally in 20 years. However, the adoption went from the tipping point to 80% in just 10 years. All the while, the US concurrently built the oil industry (distribution infrastructure), a national road infrastructure and fought WWI.
Tony touts that technology S curves are getting steeper. Adoption is happening faster and faster.
Analysts' projections are very often linear. Analysts often don't take into consideration the fact that adoptions are about systems, technology improvements working together. He cites forecasts made by the EIA (Energy Information Administration) as an example. The EIA has consistently failed with many projections.
Tony talks about technology convergence, a set of technologies that come together at the same time. He says disruptions happen from the outside. It's very rare that a company disrupts itself.
He briefly discusses ride-hailing, Uber & Lyft. The smart-phone made ride-hailing possible. In just eight years ride-hailing went from 0% to 20% of vehicle miles driven in San Francisco. He predicts that in 2020, we will realize peak new-car (globally). Car ownership will begin to decline thereafter.
Tony talks about the concept of "Market Trauma." Mainstream analysts say that EVs are "only 2% - 3% of the market, how much damage can it do?" They say "it's going to take 10 - 20 years before this takes over." A technology can disrupt the economics of an incumbent way before they have 10 - 20% of the market.
Small changes can have swift, dramatic impact in existing industries. In 2014, Tony wrote a book called "Clean Disruption of Energy & Transportation." The book focuses on these areas; batteries, electric vehicles, autonomous vehicles, on-demand transportation and solar.
In 2014, he made a predictive cost curve for lithium-ion batteries. He predicted that lithium-ion batteries would cost $100/kWh by 2023. His cost curve has actually proven to be a little conservative. Tony gives batteries storage as an example of "Market Trauma."
One example is the Tesla battery bank in Australia. The Tesla battery holds only 2% of the market capacity (ancillary services market) and yet has taken 55% market share. It has pushed down wholesale prices by 90%. Incumbent revenues have been brought down by 90%. Natural gas peaker-plants are being stranded. He also cites GE's mistaken choice to invest heavily in natural gas electrical as another example of market trauma.
Tony transitions to talking about EVs. He talks about the "gas savings” EVs enjoy. EVs are much cheaper to operate. They are up to ten times cheaper to maintain. He shows a clip of the Rivian truck doing tank turns, as an example of EVs being a better product.
EVs have a much longer life span, up to 500,000 miles, up to 2.5 times longer than IC vehicles. This is of particular interest to fleet operators. It makes total sense for fleet managers to go full EV.
He shows his cost curve for EVs. He predicts in his curve that basic 200 mile EVs will cost as little as $12,000 by 2025.
Tony predicts that next year is the EV tipping point, "for purely economic reasons." He says "it won't make any sense to buy a gas car." He predicts that every new car after 2025 will be electric. Tony cites Amazon's order of 100,000 delivery vans from Rivian, "for purely economic reasons."
Tony goes on to talk about autonomous technology. He features Waymo's autonomous ride-hailing. More than four dozen companies are investing in autonomous technology. He says "Think of EVs as computers on wheels." No one is waiting around to create autonomous technology. Only two companies will survive (only two autonomous companies).
Autonomous vehicles are safer than humans. Prediction: by 2030, we are going to be talking about taking away drivers licenses from humans. Insurance costs for human drivers will go up.
Tony brings up computing power. How quickly is the supercomputing cost curve improving? In the year 2000, the largest supercomputer on earth cost 46 million dollars and could do 1 TeraFlops (Sandia National Labs). The Apple X on iPhone released in 2019 can do 5 TeraFlops and costs about $600.
The improvements in AI are double exponential. Tony cites an AI learning to play AlphaGo and beating the world champion, then the next generation AI learning to beat the previous generation AI within days rather than months or years.
The real big disruption is in the convergence of electric vehicles, ride-hailing and autonomous vehicles. This convergence will create Transportation as a Service (TaaS). "Everyone is going electric." DIDI (the China equivalent to Uber) expects to have 1 million electric vehicles on the road by 2020.
Tony predicts that when autonomous technology goes live (approved) consumers will face the choice of buying a vehicle or using TaaS. TaaS will be up to ten times less expensive than vehicle ownership. Tony predicts that eventually, TaaS will cost less than 18 cents per mile. He says that by 2030 95% of all vehicles miles driven will be done by TaaS fleets.
By 2030 vehicle ownership will be 60% fleets and 40% personal. However, most of the miles will be driven by TaaS fleets. People will save, on average, $5,600 a year. The total US vehicle fleet will shrink by 70%. There will be fewer cars. Those fewer cars will be doing most of the driving miles.
By 2030 Taas will save the economy 1 trillion dollars per year. US disposable income will increase by 1 trillion dollars per year. The cost of travel will be only 5 cents to 10 cents per mile. This will have implications for the economy, social, health, work and other areas.
Tony predicts that oil demand will peak this year or next (2021). After this, the price of oil will eventually fall to $25 per barrel.
He talks about parking lots. He says 80% of parking will become obsolete. There will be a drastically reduced need for parking lots. That space can be re-utilized. It can be used for other things.
Tony says this disruption is not just about transportation, everything is changing. Now is the time to imagine what type of city we want in 10 years. He says that it is as if we are in 1900, we are on the cusp of the deepest, fastest, most consequential disruption in 100 years, and perhaps ever.
From the comments on the video, Tony has been saying these things for a while. It is of note, though, how closely he has come to the mark.
I question the 10 cents per mile for ride-hailing and 70% fewer automobile ownership. For the ride-hailing to be this low, electricity would have to be very cheap, and the cost of the autonomous vehicles would have to be extremely low. If an autonomous vehicle was available for $20,000 and it could go 1 million miles with minimal maintenance, then the amortized cost could be 2 - 3 cents per mile. Add to that electricity, at least 4 cents a mile. Add to that the ride-hailing service's cut and the total is going to be over 10 cents a mile.
Besides, if I could buy a million-mile EV for $20,000, why wouldn't I? It could be the last car I'd ever have to buy. The low-cost EV that makes the cheap ride-hailing possible also makes cheap automobile ownership possible, a bit of a paradox (sounds like the topic for another article). I guess we have to wait only a few years to see if Tony Seba is correct.
17 February 2020 addendum: After thinking about it a bit, I think perhaps that when Tony Seba predicts a cost of 5 - 10 cents per mile, he's referring to a customer using ride-sharing. The cost of the ride is being split among multiple people. This could work. It would mean that the revenue to the vehicle owner could still be 18 cents or more per mile. The question then arises, how comfortable would people be sharing a ride in a standard-sized vehicle? How often would people use autonomous vehicles this way? Looking at consumer behavior today, I have to question the likelihood of this happening. This would mean a departure from what is commonly being done now. I think it is a good question and it waits to be seen how it plays out.
The Future of Transportation, Tony Seba Keynote Speaker at the 2020 NC DOT Summit www.youtube.com/watch?v=