A Decade of Disruption – Autonomous Transportation as a Service

I believe 2020 is a sampling of the level of disruption we will experience this decade. There are many fundamental changes about to converge, and those changes will have a massive impact on our lives. Careers will change. Companies that have been around a long time will go bankrupt. New industries and new opportunities will present themselves.


Disruptions Coming this Decade

There are many changes happening simultaneously that are primed to take off this decade. Some of the coming disruptions include:

  • Autonomous Transportation as a Service
  • Artificial Intelligence
  • Remote Work and Remote Services
  • Climate Change
  • Pervasive high-speed internet + Cloud services
  • Battery technology / Energy storage
  • Cheaper energy
  • Augmented reality

Each of these disruptions is worth digging into; I’ve been focused on Autonomous Transportation as a Service because I believe it impacts or is impacted by every other disruption that is coming.

I’m going to attempt to break down Autonomous Transportation as a Service into all the components I see changing as a result of this technology and business model change.

Disruptions tend to start out very slow and may seem inconsequential at first. They grow exponentially fast, and it’s worth considering now what might happen in the future and how it might impact your business or career. Many of the disruptions listed above are already happening.


Autonomous Transportation as a Service

Autonomous Transportation as a Service (“TAAS” for the rest of this blog) is a business model similar to Uber, Taxi, or Hertz; you order a car for a ride or for a period of time. The difference is no human is necessary to complete the transaction. A computer driven vehicle will pick you up and go to the destination(s) you choose. These vehicles will primarily be electric because gas and hybrid vehicles won’t be able to compete with the lower energy and maintenance costs of Electric Vehicles (“EVs”), except for very specific use cases.

When you subtract a human driver from the equation, the model of car ownership fundamentally changes. Cars spend a majority of the time sitting. When autonomous cars are a reality, cars can spend significantly more time being utilized. Families will own fewer cars, if any, because it will be cheaper, more flexible, and cost less money for subscribers to ride in nicer vehicles than they currently do.

Assumptions

There are valid questions from smart people about if and when this will happen. To do a a thought experiment about what the future might look like, I need to make a few assumptions. It’s a great idea to challenge these assumptions; it may be impossible to defend them as they are educated guesses about the future. My assumptions are:

  1. Autonomous transportation is inevitable. Technology will improve to where a vehicle drives more comfortably and safer than any human.
  2. No one knows when TAAS will begin to have wide adoption, but once it begins it will grow exponentially fast.
  3. Lower cost of transportation will dramatically change consumer behavior.

Auto Manufacturing

The future for auto manufacturers is challenging to predict.

On one hand, there is huge opportunity for increased revenue and margins operating vehicles as a service instead of selling them. Having the expertise and supply chain for manufacturing autonomous electric vehicles at scale is incredibly valuable.

On the other hand, the number of vehicles the world needs will reduce dramatically. Most vehicles today have incredibly low utilization – they spend a huge amount of time sitting. With TAAS, vehicles will be utilized all day long. For Internal Combustion Engine (“ICE”) cars, low utilization made more sense because the drivetrain is designed to last 150k – 200k miles. For EVs, the motors and battery can last 500k+ miles, we will squeeze out significantly more miles from each car.

One shared TAAS car could replace 5-10 purchased cars. Therefore, a company producing 1 million TAAS cars is replacing 5 – 10 million new or used vehicle sales. This model won’t replace all car sales – of course some people will still choose to own a vehicle. The point is the manufacturing disruption isn’t 1:1 – the decrease in sales will be exponential and many manufacturers won’t survive the steep declines in sales.

I believe the path forward for existing manufacturers is to operate their own fleet of TAAS cars and trucks. If they continue to be dependent on wholesale of new cars, the pressure of reduced demand will be a massive financial challenge.

TAAS Hardware

Manufacturing of TAAS vehicles can be split into the hardware and software components.

The hardware is fairly straight forward – build a long lasting EV. The cost and even the range are somewhat irrelevant for a TAAS vehicle in the beginning – the revenue per mile will be high enough at first to make a good margin. In fact, for a couple years companies might be competitive in the TAAS market with autonomously driven Hybrid and ICE cars even with the higher per mile operating costs.

In the medium term, the critical hardware paths will be EV production capacity, battery production capacity, and production efficiency. EVs are important because the per mile cost to operate these vehicles is significantly lower than ICE or hybrid cars. Those initially high TAAS margins will get squeezed out via competition and the per mile cost to operate the vehicle will ultimately be the primary factor for profitability.

TAAS Software

The software component may be the hardest component to predict. Autonomous AI software is required to drive these cars, and this software isn’t mature yet. There isn’t even consensus on what hardware is needed in terms of sensors and cameras to have the best autonomous software driver for all conditions.

The tipping point for autonomous driving software is when it becomes as good as an excellent human driver. At that point a licensed driver will still be required, but those drivers will prefer to be driven autonomously. Autonomous miles driven with fewer interventions will rapidly make these software systems an order of magnitude safer than any human driver.

Even when the car can safely drive itself, there are still hundreds of smaller problems to solve to operate a TAAS fleet. A sample of the challenges might include:

  • Matching up cars with passengers sometimes requires a phone call between driver and passenger.
  • Usually in an Uber there are some last mile instructions like “drop me off here.” There will need to be an equivalent experience built, perhaps that’s touchscreen focused.
  • There are unmapped areas like driveways or parking lots where humans are making a random decision.

I believe TAAS companies will still need remote humans to make the service operate efficiently, and that means entire service divisions and even more software that needs developed before the car can be “driverless”.

ICE Manufacturers

Existing ICE and hybrid manufacturers are in deep shit. They need to kill off all their existing product lines, ramp up on new EV technology, disrupt their sales and distribution model, and gain software expertise all while having enough cash on hand to not go bankrupt during the transition.

In a fascinating turn of events, they may be able to do all this with the “dumb luck” strategy. If autonomous software matures at the exact moment they are ramping up their EV production, they may be able to survive the disruption by leveraging a new business model as they scale out production. They may even be able to leverage higher margins of TAAS cars with hybrid vehicles as they get their EVs ramped up.

I’m skeptical of this working out. It’s a recurring theme throughout history that unless a company is willing to compete against their own existing sales, they do not survive major disruptions. ICE manufacturers are not attempting to disrupt their own companies, they are trying to “ease in” to the future.

Sales are going to crater for ICE cars for two reasons: 1) the purchase price for an EV with superior specs will be lower than an ICE car and 2) the TAAS model will move a significant percentage of consumers to a shared car model. Once the sales start to decline, the overhead for producing ICE cars will increase, which will further accelerate the decline. While the manufacturing cost per new ICE car is increasing, the price for used ICE cars will crater as people decide car sharing will save them a lot of money and they abandon used ICE cars. The decline in the price of used cars will put even more downward pressure on new ICE sales.

I shake my head when ICE manufacturers talk about their ICE sales projections 10 years from now; I believe new vehicle ICE sales 10 years from now is “quickly approaching zero”.

TAAS Focused Manufacturers

First mover advantage will be critical for TAAS services. In any local market there will only be a handful of winners – these fleet networks need a critical mass of users or specialized service offerings to be profitable. We will probably end up with 5-10 global companies who are operating autonomous fleets, with perhaps 5 operating in any local market. Companies who can gain market share first will have a huge advantage.

Autonomous cars that are not part of a shared network will still need TAAS-like services. For example, if a personally owned car is going somewhere without a human driver, someone will still need to be responsible for the vehicle in case of a problem. This responsibility will probably require a service offering from a TAAS provider.

Government Regulations

I don’t see government regulations restricting autonomous cars being that big of a hurdle. As long as the safety is demonstrated, there are already legal and insurance remedies for accidents. An abundance of data will be available to demonstrate better safety than human drivers.

A bigger hurdle during the initial rollout of TAAS could be media stories that miss the forest for the trees. Autonomous cars will have accidents, the accidents may look dumb to humans, and media companies will write about whether we should trust computers to drive us. Hopefully our legislators will make data-driven decisions and not “this single data point looks bad” decisions. The goal should be to significantly reduce accident rates and injuries, not demand zero accidents.

The biggest regulatory risk is for ICE and hybrid cars around carbon emissions. Governments are already implementing stricter emission standards, both to combat climate change and to get cleaner air in our cities. I could see governments demanding TAAS cars either be EVs or face high per mile carbon taxes, effectively pushing those types of vehicles out of the market. For example, to avoid regulatory fines, in Europe ICE manufacturers are currently paying their EV manufacturer competitors cash money! Once governments see TAAS as a viable option they are going to demand the switch to zero emissions.


Automotive Services

Industries that are downstream from auto manufacturers are also going to see significant disruption. The business model is going to change, and existing players are going to be left out or purchased for a fraction of the current value.

Sales and Distribution

Independent dealerships are not going to survive with their existing business model. Car sales will drop off a cliff, and the revenue related to warranty and maintenance will decline proportionally. EVs also have lower maintenance costs, causing further struggles with operating a profitable service center.

The existing dealership infrastructure does have value – dealership service centers could be repurposed to clean, wash, charge, and maintain car fleets. Dealerships also have a local presence which is an asset in maintaining a huge distributed fleet of vehicles.

Dealers are so tightly tied into ICE manufacturers and vehicles that it’ll be tough to transition. It may be wholly dependent on whether their supplier has a good TAAS strategy. Some dealership networks are going under, and at that point successful TAAS companies will buy up the real estate at a discount.

Used car dealers will also struggle. Over time, it will be cheaper to subscribe to TAAS than the gas, maintenance, and insurance costs of a non-autonomous ICE vehicle. Used cars will plummet in value. Consumers will buy up the luxury cars at a more accessible price point. Older, cheaper ICE cars may fall to the value of scrap metal.

Taxis

Traditional taxis are probably the first industry to get wiped out. They are going to zero. Their existing assets don’t have value in a TAAS world.

Ride Share Companies

Ride share companies can only survive by changing their business model.

Ride share companies currently have significant brand awareness, user adoption, and some technology assets for matching cars to users. However, if a company is not manufacturing EVs, they are not in control of the critical component for TAAS. Auto manufacturers will end up with the leverage, because it takes years to ramp up vehicle manufacturing at scale.

Ride share companies are headed towards mergers with EV manufactures.

Long Term Rental Companies

Rental car companies probably won’t survive. A lot of this market will be displaced by the ride sharing model. Even for longer trips, it’s still effectively the the same business model as ride sharing. I don’t see why manufacturers would sell cars to third parties when they can run their own service and earn significantly better margins. Like ride sharing companies, acquisitions could make sense for this sector.

Similar to car dealerships, rental car companies have value in their real estate holdings and car servicing operations . TAAS companies will purchase rental car companies for their real estate and service centers, but the value of the existing rental car fleet business is headed towards zero.

Ride Share and Taxi Drivers

Professional drivers are mostly going away, except for specialty scenarios. Most will have to find new careers. Some in-vehicle personal services like bus “monitors” or helping people with mobility issues will remain as careers.

New services will crop up around remote assistance for passengers. Perhaps a passenger needs help finding a location, or some other issue. Video calling and manual car navigation staffed by locally knowledgeable people will provide a value-add service for riders.

Delivery

Delivery disruptions are probably longer term. Significant “last mile” hurdles exist for delivery and keeping humans in the mix may be more cost effective.

Some specialty vehicles will be produced around delivery. Perhaps your pizza is delivered by a motorcycle sized autonomous vehicle, but you have to go to the street to retrieve it.

Maps and Navigation

Although maps have been around a long time, there are still times where humans are solving the details. Navigation needs more investment and improved accuracy.


Add-on Services

Service and Maintenance

The service requirements for autonomous cars will go up compared to existing cars. They will need to be constantly cleaned, washed, charged, and inspected. Jobs in this sector will increase, and existing infrastructure (e.g. a car wash) will be repurposed for taking care of fleet cars.

Dispatch services and local service technicians will be needed to triage vehicles with mechanical issues.

Independent mechanics will see a decline as EVs don’t require as much maintenance as ICE cars. However, the existing ICE fleet is massive. If manufactures and dealers go bankrupt, independent mechanics may actually see an increase in business for a while. But in the long term, this will be a declining business.

For a time there will be a decent size retrofit business for adding autonomous driving features to existing cars. Companies experienced in electronics retrofits, like stereo installers, could be in a good position to capitalize on the new technology.

Insurance

TAAS companies will be self-insured, eliminating a large portion of revenue for existing auto insurance companies.

Even if they are not TAAS cars, autonomous cars in general will be safer than human driven cars. Cars primarily driven in an autonomous fashion will be cheaper to insure. Even when not operating as part of a TAAS fleet, car manufacturers will offer a significant discount for miles driven in an autonomous mode. In fact, the insurance for driving autonomously may be so cheap that the savings will offset the cost of the autonomous software. Existing auto insurers won’t have the data necessary to properly price autonomous driving.

Auto insurance becomes a niche market for legacy cars and specialty vehicles such as motorcycles. Existing providers will go bankrupt and consolidate as the auto insurance market disappears.

Traffic Enforcement

Police departments will have less revenue from writing tickets for infractions. In the short term, human drivers will feel the pain of this disproportionately; an autonomous vehicle should rarely get a ticket. Long term, cities need to come up with new revenue streams.

Towing companies will see significant contraction. TAAS cars will be better maintained and won’t need to park illegally.


Automotive Competitors

Public Transportation

Traditional fixed bus routes may be replaced with smarter routes in smaller vehicles. If passengers can enter their destination and a range of arrival times, a TAAS network can plan out a mix of vehicles to optimize the commute time for everyone.

Shared transportation like trains, subways, and air transportation will be enhanced by a TAAS network. Passengers will find it easier to get from hub stations to their final destination.

Freight

Trucking and rail services will see significant disruption from autonomous trucks. It might be possible to double or triple the utilization of an autonomous semi, plus save significantly on fuel costs. TAAS truck freight might be cheaper than rail, putting a strain on the entire rail business.


Energy

Clean, renewable energy has its own disruption story. EVs will leverage and accelerate those disruptions.

Energy Arbitrage

A fleet of TAAS EVs can help with an additional function called energy arbitrage. Renewable energy is sporadic – the sun doesn’t always shine, and the wind doesn’t always blow. Stationary batteries help distribute energy when it’s required; EVs can also play a part as a form of stationary battery. At non-peak transportation times vehicles could choose either lend energy, charge batteries, or neither. Managing the vehicle battery levels intelligently adds value to the fleet through cost savings and selling energy back at peak rates.

Oil and Gas

Oil companies (and oil reliant countries) are also in deep shit. With EVs in general, oil demand will drop slowly over time. With TAAS EVs, oil demand will fall off a cliff.

In the short term, the reduction in demand will push oil prices down. However, lower prices aren’t sustainable when they go lower than the cost of production. Eventually, unprofitable oil wells will be closed, and oil supply will go down. As supply is reduced and oil companies go bankrupt, the economics will change and prices will increase.

National governments and citizens are concerned with climate change and are going to increasingly tax polluting vehicles. Cities where citizens demand improved air quality are also going to change the economics for operating ICE and hybrid cars.

Local gas stations are going to be harder to find as demand goes down, making driving an ICE car even more inconvenient. With TAAS, fleets will be charged at centralized servicing locations. The reduced customer foot traffic kills high margin Slurpee sales and many gas stations go out of business. Neighborhood stores may be able to survive if they have a local consumer base.

Highway and interstate gas stations may survive by adding quick charging stations, but it seems like many gas companies are not reacting fast enough. New EV infrastructure is being built at competitive locations; there may be no desire to put high speed charging infrastructure at existing gas stations by the time they see demand falling.

EV Charging

TAAS networks will not need super fast charging to succeed. Fast charging is more important for a long-range trip than for operating a TAAS fleet. TAAS EVs will start with a full battery pack every day by being charged overnight when people are mostly sleeping. During the day, there are morning, lunch, and evening peak transportation times. EVs will have enough energy to manage peak demand, then can be recalled for charging and cleaning when demand is lower.

Energy Prices

The cost of solar energy continues to decrease dramatically every year, which will be a disruption in itself. EVs are uniquely positioned to take advantage of these savings, and TAAS EVs will see operating costs drop even further over time. With energy arbitrage, TAAS EVs providers will be paying the lowest possible rates for electricity.


Construction

Changes in transportation will kick off a global construction boom, as cities are able to rethink their existing use of space previously dedicated to individually owned cars.

Parking

Cities will reclaim street parking spaces. Some street parking will be dedicated as pick up / drop off lanes or additional traffic lanes. It will also be repurposed away from vehicles.

Public and private parking lots and garages will stay around a bit longer, but will slowly be repurposed as the demand goes down. This will give places like hotels and retailers new design options for the reclaimed space.

My hope is that cities and businesses use this as an opportunity to make more space for pedestrians, cyclists, outdoor restaurants, and green spaces.

Infrastructure

In busier settings like airports or urban centers, new infrastructure will be needed to help passengers get matched up with their vehicle.

Although autonomous cars will be able to read human signs to drive on existing roads, it will be desirable over time to redo infrastructure to be purpose built for autonomous drivers. For example, on a road that allows only autonomous cars we could introduce the concept of “4-way yield” instead of “4-way stop”.

I believe city and building design will change in more fundamental ways that I cannot currently imagine, and the related construction activity will be massive. Every component of existing infrastructure needs to be rethought.

Home Design

Home design is still going to be heavily dependent on climate. In cold or rainy ares, people want to be protected from the elements when getting in their car.

In general, houses and apartments will focus less on parking vehicles and more on efficient “pass-through” of vehicles. This will change how housing is built – less space needs to be allocated towards garages and driveways.

Many existing single family home garages will be repurposed. Home office space or home theater space could be logical uses for existing garage space.


Government

It’s impossible to predict how governments will respond to this technology; here’s what I think they should do.

Autonomous Regulation

Governments can come up with some simple autonomous driving metric requirements prior to allowing the fleet to operate driverless. Once the safety of passengers and pedestrians is proven, it can be opened up and continued to be measured. Example metrics might be:

  • Number of human interventions per million miles has to be at a minimum level.
  • Accidents per million miles needs to be at least twice as safe as expert level human drivers.

Once the metrics are met, the car can be driven (in that specific geographical area) without a human driver or without the human driver required to pay attention.

Taxes

Governments should come up with per mile taxes for autonomous rides. Part of this money should be allocated towards infrastructure and part should go towards helping people who’ve been displaced due to the technology changes.

Empty vehicles should be taxed even more, especially during peak traffic times. TAAS operators and individuals should be financially incentivized to optimize driving times and not clog roads with empty vehicles.

Freight trucks should also be incentivized against operating during peak traffic times. Trucks can operate overnight and daytime when traffic is low, charging during morning and evening commute times.

Education and Job Disruption Programs

Disruptions tend to bring new opportunities for society as a whole, but are simultaneously harmful to individuals. We need education programs funded by the disrupting technology that help individual workers start new careers.


Grab Bag

Small Towns

Smaller towns could be some of the biggest beneficiaries of TAAS.

Partly this depends on what happens long-term with knowledge workers working from home. Will COVID-19 cause a long-term trend away from big offices in cities, or will things go back to how they were? If people only need to commute once a week or once a month, they might be willing to live in a cheaper community further outside larger cities. TAAS would make this more plausible, as the commute itself could be productive work time. People might be willing to live further from cities if they can be transported there stress-free.

Smaller towns also instantly gain the equivalent of a public transportation network for their citizens making it easier to get around.

Traffic

It’s hard to predict what happens with traffic.

On the plus side, there will be fewer accidents and a smoother flow of traffic. Certain types of traffic can be more intelligently deployed, such as freight traffic limited to certain hours. Also, commuters might be willing to utilize more shared transportation like mini-buses if the route is more direct and they can occupy HOV lanes.

On the minus side, people will be willing to drive further distances if they’re not responsible for operating the vehicle. People might live and commute further from city centers. Additional commuting miles could mean more cars on the road at any given time. If transportation is cheaper, commuters might also move away from trains or buses.

Smart Phones

To be part of a smart transportation network, users will need at least a basic Internet connected smart phone. For some users, the transportation network itself may provide a device as part of the service costs.


Summary

I believe the TAAS disruption is going to be massive. Governments, companies, and individuals should be thinking through the future impact and starting to plan now for the future. There will be a lot of opportunities and a lot of heartbreak as the world changes.

In 2020 it’s been hard enough to make 1-day decisions; most of us are not thinking about 5-year decisions. However, disruptions don’t wait for us to get properly prepared.

One of the biggest disruptions we’ve seen since the Internet is around the corner.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s