In 2011 EIA (US Energy Information Administration) predicted that there would be 200K battery electric cars sold in
2035 in US [
source][
source] (in 2011 Tesla was already producing and selling the original Roadster). That prediction did not age well. Instead, in 2019 there were 330K battery electric cars sold in the US [
source].
"Why did the smartphone happen in 2007? Not in 2005 or 2009? "
The concept of technological convergence is important to understand because it is often underestimated. Tony Seba ask a question in one of his presentations:
Why did the smartphone happen in 2007? Not in 2005 or 2009?
[source]. 2007 was the point when all technologies that made it possible converged: computing power, cellular Internet, data storage, touch screen combined into a $600 device, a smartphone. Many companies could have created a smartphone: Nokia, Motorola etc. but such monumental disruptions happen from the outside not inside (see
Innovator's Dilemma book). Subsequently Uber and Lyft were made possible by smartphones few years later. And 8 years later Uber has more daily bookings than the whole taxi 🚖 industry [
source].
25% of all vehicle miles travelled in a U.S. city are now booked on demand [
source]. Sales of passenger cars are starting to be affected by this. Peak of new car sales is coming, in fact it looks like 2018 was peak of new gasoline car sales in the U.S. [
source]
New technology can have swift and dramatic impact early on even before it takes significant share of the market. Even though solar, wind and batteries are still small percentage of the market one can already see signs in the weak outlook for General Electric that continued to bet on thermal energy generation (and not investing into grid battery storage). What will drive disruption of energy and transportation further? Cost curve of Lithium-Ion battery technology. Wright's Law continues to drive down the costs of Lithium-Ion batteries [
source].
In 2018 Tesla installed a Powerpack (a large electrical grid battery) in Australia to help smooth out power demand spikes that previously necessitated use of thermal peaker plants. With 2% overall market share in capacity the installed battery brought down incumbents' revenues by 90% and paid for itself within the first two years of operation [
source]. Thermal power (fossil fuel) peaker plants are now being displaced by battery grid storage. NY state just announced a massive battery project that will replace up to 16 natural gas-fired turbines at a sprawling plant that supplies about a fifth of New York City’s power. “This facility will displace energy produced from fossil plants during peak periods, resulting in cleaner air and reduced carbon emissions,” said PSC chairman John Rhodes. [
source]. GE planed for demand for gas and coal powered plants to continue growing. Their stock is now down by 70% driven by collapse in new thermal power construction market.
Lifetime, on average, of an EV is 2.5x to 5x that of an ICE car. It makes especially sense for fleets to go electric.
GE has wind 💨 energy generation sub-division, yet they expected those to not affect the market. Can we draw parallels to auto industry, how real is the battery electric vehicle market disruption? Battery Electric Vehicles (BEVs) are 10x cheaper than conventional ICE (internal combustion engine) vehicle when measured in life-time cost per mile. Maintenance is much reduced vs. an ICE car due to much lower complexity of the drivetrain. Lifetime, on average, of an EV is 2.5x to 5x that of an ICE car.
Price-wise, parity with a median passenger ICE car has been reached in 2020 ($35K) in US and the cost will continue to come down. By 2025 we will be able to buy a $15K EV with 200+ miles of range. By then every new car will be electric for purely economic reasons. In 2019 Amazon ordered 100K battery electric delivery vans from Rivian [
source], solely for economic reasons. UPS, Fedex will follow suit. In fact, UPS just placed an order for 10K electric vans [
source].
Autonomous technologies are the new computer Operating System; it's a winner take all. Waymo has ordered 62K self-driving minivans to be delivered this year [
source].
Are autonomous vehicles safe? According to Tesla data, they are 8x safer today (when driven on Autopilot) than in human drivers' hands [
source]. By 2030s we will be discussing taking away ability for humans to drive cars, or making it very expensive. With 40% year over year improvements autonomous driving can be 100x safer than a human driver by 2026.
Jim Keller of Intel on autonomous driving: "It's a computational problem, understanding ballistics, topography. It's not an intrinsically hard problem."
On a
recent podcast, AI researcher Lex Fridman asked Jim Keller whether autonomous driving can be solved on the timescale of the next few years. Jim's response: "definitely, it's a computational problem, understanding ballistics, topography". It's not an intrinsically hard problem.
1 Teraflops of compute in 2000 cost $15M. In 2016 8 Teraflops of compute costs $600. Today 25 companies are developing Level-5 autonomy (full self-driving) technology. Anything on wheels is going to become autonomous.
Insurance costs are going to go up for human drivers. We could be saving 1000 lives every year just in New York State.
Last mile delivery will be solved by remote drivers:
So where’s the big disruption? It's in the convergence of: autonomous, BEV and on demand transportation. Cars are the second largest expenditure after home, yet we park them 96% of the time and only use them 4% of the time. The cost per mile of electric, autonomous, on demand transportation is 10% when compared to conventional fossil fuel car ownership.
Today average car sits idle
of the time.
Cars are being utilized
of the time.
Effective cost of an on-demand, autonomous EV will be
of the costs of conventional ownership of fossil fuel car
As an example, a fleet of 1M EVs can drive 80% of all vehicle miles in New York State where there was a total of 5M cars registered in 2019. For scale there is a total of 276M vehicles registered in the US. What happens to auto industry when we essentially cut it to 20-30% of that number to meet the transportation needs in 2030? Transportation as a Service is going to be a major disruption. On demand, autonomous, electric, owned by fleets.
Let’s assume that by 2025 autonomous driving is approved by regulators widely, then the cost per mile drops by 90% overnight. Would we continue to spend $50K on a new car ($35K purchase + $15K in fuel costs and maintenance spread over 5 years) or a $100 monthly subscription to Uber, Lyft etc. makes more sense? By 2030 80% of all vehicle travel in America is going to be on demand, autonomous and electric. Fleet vehicles will cover 80% of all driven miles.
The implications are going to be astounding. People will save $5-10K per year by not owning a car. Car ownership will drop by 70%.
Transportation is going to be essentially free.
It will result in boost of disposable income by $1T [
source]. Additionally, productivity will increase due to freeing up of time for those driving today. Transportation is going to be essentially free.
Vehicles are going to be redesigned to look more like living rooms. Transportation CO2 emissions will decrease by 90%. Oil demand will peak this year.
We will need to envision what to do with all this reclaimed space.
By 2030 global oil consumption will drop by 30% from the current levels [
source]. Additionally, some of our cities have as much as 40% of land dedicated to parking [
source]. Majority of that parking space is not going to be needed anymore. That means there will be more space for parks, residential areas, offices, etc. We will need to envision what to do with this reclaimed space.
