Why Robotaxis Will Struggle To Compete

Support CleanTechnica’s work through a Substack subscription or on Stripe.

First of all, compared to taxis and app-based ride-hailing vehicles (Uber and Lyft vehicles), robotaxis may end up competing very well and driving those human-driven options out of the market. However, there’s a huge difference between that limited market and the normal passenger car market. Many robotaxi enthusiasts expect robotaxis will largely replace those normal passenger cars, which would be a truly disruptive market revolution.

Recent comments from longtime readers “Matthew2312” and “eveee” brought this topic to mind again. Here’s the extended comment from Matthew2312:

“Well we already have TAAS [Transportation as a Service] widely available in every market, right? That is literally what the taxi/ride hailing business is. Even more than that, we have courier services (including UberCourier) and delivery services which are also TAAS but where your stuff rides and not you. Then we have vehicle by the hour and vehicle by the day services from various rental car companies (and the failed business roster is littered with grab-and-go vehicle rentals that were going to change the world). IOW TAAS is here, it works, no automation required, we know the size of the market and its growth.

“The only argument for why it might be bigger is the argument that automation will reduce the cost and that there is a high elasticity of demand, right? But is there any evidence that is the case? The cost of owning and operating a vehicle you purchase new and drive for five years, depending on various assumptions, is $0.75 – $1.00 per mile. So if you cut taxi cost by 50% vs the current Uber/Waymo rate, you would be at break even. But who would rather give up a car for a taxi at break even? Especially since this is the new car purchaser who has the money (they buy all new vehicles and determine the fleet).

“Okay so then you have to get to something more like 50–70 cents a mile to even start the ‘I’m selling my car’ conversation. Can you run a robotaxi for that? Well the human driver only saves you $0.50–$0.70 of the $1.50 – $2.00 current cost. So, no, you cannot. You can conceivably get to an operating cost of $0.40 (a lot of magic there), but then you need to actually pay the company overhead. A profitable robotaxi company still needs to charge $0.80 per mile or more even with so-far-unattainable operating costs. This is why an internal Tesla study said the robotaxi business was not attractive (Elon rejected).

“The obvious problem: autonomy is not limited to taxis.

“Now think about this for a second. If autonomy is available, then a new car-buying household — which will tend to have multiple vehicles — will have personal autonomy. That means it is much easier for them to share one vehicle (send it to pick up someone from work). That would lower their own operating cost from $0.70 to under $0.50 perhaps supplemented by a taxi.

“We can keep pulling on this thread, but the reality is that (a) robotaxis at current or slightly lower costs will displace human drivers because the service is better and that is around 350,000 vehicles by 2035 (b) they will incrementally grow the business if they can get prices down below $1.00 per mile, which would goose it to 500,000 or so (c) large-scale TAAS migration will not happen in the US even at $0.70 per mile. There ARE markets where things may be different (crowded Asian markets in particular), but in the US, this is sub-50,000 per year new robotaxis on the road.”

Indeed. Also, note that a small sedan can come at a cost of just ~$0.55/mile. So, if your focus is lowering costs, that makes the robotaxi case even much more challenging.

If your focus is convenience, as the commenter above points out, a robotaxi is not going to be more convenient than a self-driving car of your own, or even a normal car of your own for many people. When you own a car, you go outside or to the garage where it’s parked whenever you want and then immediately drive to where you want to go. With a robotaxi, you have to request the ride and then wait however long the wait is for the robotaxi to arrive. And you can’t leave stuff in the car as a mobile storage space or extra room.

The other problem with all of this is that when you have your own car, you drive from point A (where you are) to point B (where you want to go) to point C (where you want to go after that) to point D (back home); whereas when you use a robotaxi, instead of A–B–C–D, you get the following:

• It drives from point Z to point A,
• Then from point A to point B,
• Then from point B to point Y,
• Then from point X to point C,
• Then from point C to point D,
• Then from point D to point W.

You see all of the extra driving there? That extra driving isn’t free. So, even if a robotaxi is hyper efficient, that extra driving adds up to more cost. Imagine the Cybercab costs just $0.50/mile, and your normal car costs $0.77/mile, but the Cybercab has to drive twice as much to provide the same trips. That makes the Cybercab considerably more expensive.

Here’s how “eveee” responded to “Matthew2312” the other day: “Right Matthew. The user costs of TAAS run about $2/mile in cities US. Private owner costs are about $0.75/mile. Economically, it makes no sense to switch to TAAS even if driver costs are eliminated.

“There is another reason why increases in TAAS will not work (even if it is a privately owned vehicle). Traffic congestion. TAAS has already worsened traffic.

“It may be that market push and advertising can push consumers into some taxis, but claims like those of Tony Seba are out of touch.

“For some reason there is way too much focus on products and objects and way too little focus on systems, cities, transport infrastructure, moving people.

“We might as well be discussing the future of Segway for all road transport because it has superior door to door function.”

Indeed.

I’ll confess — I was enthralled by the robotaxi concept at one point (7 to 10 years ago). However, these kinds of points become more obvious once you dive into the topic. And it’s hard to figure out, given everything presented above, how robotaxis are supposed to take over the market. Your thoughts?