We are at interesting cross-road right now with industrial automation principles starting to leap into the consumer space. But what does it mean, how will it impact, and what will the ‘new math’ look like?

I’m going to drive the discussion focused on cars, but cars is only an entry point, and you can certainly take the logic to other spaces.

And forgive the awkwardness of my ‘autonomous car study‘ – it is by no means comprehensive, but it should be enough to get the brain moving in some new directions.

Premise

The premise is we will all soon be able to use self-driving cars. They drive themselves. They will take care of themselves. You no longer drive them. You decide where to go and the car takes you there. Full stop.

Impact

The impact goes beyond humans driving. If you no longer need to drive yourself there, do you really need a car to be ready when you need it? Why can’t you simply call a car when you need it instead of having it sit idle in your driveway? If the car can self drive you somewhere, why can’t the car self-drive to pick you up as well? In this paradigm potentially car ownership is less likely, and car ownership might remain with the car makers, or maybe some middle groups of fleet owners.

Agree, or agree to disagree – either way is fine, as we’re really here to talk about the autonomous math…

Current Car Math

To start on the math part, let’s figure out what it currently costs a vehicle to get anywhere.

We’ll cover this a few different ways… all interesting and all good angles to do some mental stretching.

According to the IRS, the 2024 mileage allowance rate is $.67/mile. So if you drive a car and are going to claim it as an expense, Uncle Sam believes you should receive $.67/mile,

Boston taxi rates are a blend of mileage and idle time, with First 1/7 Mile: $2.60. Each 1/7 Mile thereafter .40. Toll. $28/hour of idle time.

New York City taxi rates are $3.00 initial charge plus 70 cents per 1/5 mile when traveling above 12mph or per 60 seconds in slow traffic or when the vehicle is stopped.

Uber rates are a combination of base + fee + time of day – you know the price before you get in. But conventional thinking is that this solution is cheaper and more convenient than your current taxi experience.

In the various livery scenarios you are not the car owner you are so you not carrying any wear-and-tear, insurance or garage fees. And of course you are not directly paying for fuel.

Let’s say you do own a car… and if so, and we’ll call you average that means you are driving 14.263 miles per year.

But let’s not quibble over miles and calculations and instead lets do this part as easily as possible courtesy of AAA, so all-in, all-up in 2023 the cost of owning a car was $12,182 – if you really want to pursue a customized number based on your make, model and location here is more granular cost of car ownership calculator.

Okay – so for our purposes, if you own a car you are looking at around $12K/year and that doesn’t include any kind of garage fees.

Autonomous Car Math

So here it gets interesting.

Now let’s drill in a bit more on the short-term rental side, (taxis) – and I get that we’re fishing for data, but the New York Times feel pretty reputable to me, so here goes – average New York Cit taxi fare is $19.62.

To start on the math part, let’s build on the $19.62 amount (and I get that you might want to think about a different number)

Here is a link to my Autonomous Car Cost Calculator spreadsheet – feel free to alter it however you’d like – make things cost more and less – my feelings will not be hurt. : )

And we should factor in some additional elements…

In the autonomous scenario you do not need to drive, so no need for a driver. For a car owner this is a non-factor as you still need to be transported, but for a livery (someone else driving a passenger(s) somewhere) this is a huge cost that no longer exists. No hourly rates. No health insurance. No food. No stopping to take breaks.

We’ll also likely see economics of ownership at scale. And if we no longer own the cars, but rather the car companies do (or some other big groups) we can assume they’ll continue to find efficiencies around other costs. You can imagine buying a fleet of cars is going to cost less than buying all those cars one at a time. How about if you don’t even bother buying a fleet but rather you are the car manufacturing – suddenly all those pass throughs and markups disappear…

Other car costs will also come down – we’ll no longer be focused on mass customization – do you really care about all the features in your taxi? I didn’t think so. Uber and others have figured this out to some extent through a very simplified tiering system (Uber X, XL and Premier). Think about it. In this world there might only be 3 colors of cars per make. Ugly yellow. Mid-range candy red, and expensive black.

So less features, and less colors.

And of course if the car is driving itself no steering wheel and pedal. So more savings.

So what will it really cost?

Depends who you ask… but according to Elon Musk, Tesla fully autonomous cars will retail for around $25K. Other analysts peg this figure around $200K. Who is right I really don’t know but I’d have a hard time arguing against the Tesla group.

Autonomous technology is still early days, but there is certainly a race happening right now that involves hardware (physical cars) and software (a mix of driving and logistics technologies) – who wins I don’t really know, but as this space scales and matures the costs will eventually costs here come down.

So to net it out, and the high end an autonomous car according to my murky math can generate $220K and and autonomous car can cost anywhere between $25K and $200K, so even at the high end the potential take home can be significant over a multi-year period if there space matures, and my murky math holds.

Consumer Focused Autonomous Car Math

Consumer costs should also come down – anything below 12K/year is a ‘win’ for the consumer, so for the urban (high density living) consumer this should be a net positive. For everyone else it might make more sense to continue to be a car owner, and possibly even rent your own vehicle out when not in use…

So you might lose some features, but you’ll also lose some headaches around car ownership.

As you’ll see in the above sheet, I did a little cost estimate on a ride per day, 2 rides per day… and that is not factoring in for efficiencies like autonomous car usage rates driving the fare costs way down, enhancements around logistics and basic ideas like ride-sharing (carpooling) and other grid management gains. The average cost per ride should also come down rather significantly.

On the other side, if you really do need to own a vehicle (back to the autonomous car math…) with an autonomous vehicle just think about all the other factors that will be more favorable to the passengers, particularly those pesky details that used to have to fall onto the shoulders of the driver…

And maybe, just maybe when you car isn’t driving you, you might want to put it out on the street and let it earn for you…

What’s Next

So now that we’ve ‘solved’ the world of autonomous cars (yes, I know we just scratched the surface…) , I’d suggest you continue to carry the argument forward by removing the word ‘cars’ and applying to other domains and also more broadly across entire economies.

  • Ownership of physical items versus short-term needs based rentals.
  • How services are rendered and ultimately who is the collection entity for these services in autonomous environments could be drastically transformed.
  • And of course the role of people. Both from an income and an experience perspective will look very different.

The implications are profound!

Thumbnail image of an autonomous car at an autonomous car wash rendered by Google Gemini