Will Elon and Tesla Prove the Doubters Wrong on the Robotic Taxi Master Plan

Elon described costs and details of the master plan for a robotaxi fleet and service at Autonomy Day. Many are criticizing the full self-driving plan and do not believe it will be achieved or will be far, far later. The future of Tesla and the world will be massively impacted if this master plan succeeds or if it does not.

If it succeeds then Tesla makes billions. 1 million robotaxis making $30,000 per car per year for owners and similar amounts for Tesla would be $30 billion per year.

If would also mean about 50 billion robotaxi miles per year within a couple of years of successful full serlf-driving. Uber provided 26 billion miles of ridesharing in 2018.

In the USA, 3.22 trillion miles are driven on roads and there is about 12 trillion miles driven in the world.

Tesla makes their in-house chip for full self-driving. In 2020, they expect to have 1 million cars on the road with the hardware necessary for full self-driving and which could become robotaxis. They believe they will have the most profitable autonomous taxi on the market. Uber and Lyft lose a lot of money on ridesharing. Elon Musk says that a robotaxi using a full self-driving taxi will only cost 18 cents per mile.

The average Tesla car is parked for 22 hours per day. In 2020, owners will be able to use the Tesla app to their car to pick up and drop off passengers autonomously, earning an estimated 65 cents per mile in fares.

By Tesla’s estimates, owners might be able to earn $30,000 in gross revenue from their cars per year. This would be more than $300,000 in revenue over the 11-year lifespan of an average car.

Elon is saying Tesla cars will last 1 million miles. Tesla owners would need to give nearly 500,000 miles or ride at 65 cents per mile for Tesla car owners. Tesla would make more from their share of the robotaxi business and Tesla would have post-lease vehicles dedicated to the robotaxi fleet.

Written By Brian Wang, nextbigfuture.com

25 thoughts on “Will Elon and Tesla Prove the Doubters Wrong on the Robotic Taxi Master Plan”

  1. A big stick with tar-soaked rags and a flame at the top. Or maybe more advanced like the Stature of Liberty has or they use in the Olympics.

    That thing with a battery is a flashlight…or if it is crazy big and bright…maybe a spotlight. I use a head lantern. Which is a small LED flashlight with an elastic band. Though it is often called a headlight/headlamp which can lead to confusion as that is the same words used for the lights that illuminate the road in front of the car. I think the Chinese started selling them and “headlight” sounded fine to them, not realizing that that was already in use and meant something else, and that we already had a term for a light on your head as well. Now they are using “headlamp” which may be in the process of loosing it previous meaning.

  2. Sorry, but this looks like operator error. I don’t think Autopilot has the ability to stop at stop signs yet, or even to follow a planned route (other than taking itself off the highway using an off-ramp). The driver may have put it on lane-following and let himself get distracted, missing the stop sign and T-intersection.

    Of course, one could say that Autopilot encourages that sort of bad behavior. I personally wouldn’t trust Tesla Autopilot except maybe on roads I’ve driven several times since the last software update – because I’d be worried that while learning new stuff, it might regress on some edge cases it handled OK before.

    Tesla has about 500,000 cars on the road, which would mean around 5 to 7 billion miles driven a year. At the US rate of about 11 deaths per billion miles driven, Teslas ‘should’ be accounting for between 50 and 80 deaths a year. And those cars have totalled about 10 billion miles driven, so that’d be 110 deaths total so far.

    So maybe we should stop being so shocked every time someone dies in an accident involving a Tesla?

  3. No evidence that the autopilot was driving, but that’s just the problem really. Every time a tesla crashes it’s interpreted as a self-driving accident until shown otherwise.

    And they should have never, EVER called it “autopilot” until it was actually ready to drive unsupervised.

  4. Not all transfers are created equal. E.g. going from one flight to another with a 2 hour layover, versus going from a flight to a taxi – versus going from a taxi to the elevator inside a building.

    What I’m proposing is a lot more like the latter. A taxi picks you up at your door and drops you off right beside a van that has a reserved seat waiting for you, likely under a shelter so you don’t even need to deal with rain or snow during the transfer.

  5. Yes, using a car as a storage locker is convenient. But if the alternative is “make an extra $1K-$2K a month”, maybe you’d be willing to be sure to take out anything you might need that day, and occasionally get cold or wet for the 10 steps from the steak house door to your nice warm Tesla as it picks you up…

    Or the service that takes your stuff out and later puts it back (after cleaning the car) could pull your coat from your locker and send it over to you. Maybe you get one “oops, forgot something, please send it” free each month as part of the deal.

  6. I think the issue is that people WANT to leave stuff in cars.

    My car has

    • An umbrella
    • A torch
    • The remote control for my garage door
    • Shopping bags
    • A hat
    • A first aid kit
    • A semi-respectable jacket (cold months only)

    Those are all things that I want to have available, but not carry around with me all day. I don’t want to carry a coat all day, but if someone says at 4:30 pm “hey, let’s go to that new steak house” I want to be able to just go without freezing when the sea breeze picks up after 10.
    I might only use the coat, or the hat or umbrella once a month or less, but keeping it in my car is costless.

  7. Why do companies sell shares and then pay out dividends? If they knew the dividends would earn money why not keep them?
    I agree, the basic argument disproves a lot of things that happen in real life.

  8. It’s like those bitcoin mining ASIC’s. If they actually could make money, why would they sell it to you?

    By that argument, why do shovel companies even sell their own shovels? They should do all the digging themselves and charge people for it.

    Not that I necessarily disagree with you, but its a lousy argument.

  9. One solution to heavy traffic might be to have a self-driving taxi pick you up at home (or the office) and take you to a nearby ride-combining center (generally in the direction you’d already be going). There you’d be dropped off right next to a waiting self-driven van with other passengers who can all be dropped off without a lot of side-trips. So your commute is generally only marginally longer at worst, and usually shorter due to use of car-pool lanes and reduced overall traffic.

    You’d have a variety of options – individual ‘work compartments’ versus cheapest possible ride, how much conversation you want, food allowed vs drinks-only, etc – the sort of scheduling stuff computers can do pretty easily and give you options on.

  10. I thought of a possible solution to the “leave stuff in my car” issue: The Tesla taxi service would have to provide a cleaning service. So when a car comes ‘on shift’, it could go in for a quick inspection, and anything you left in the car could be pulled out and stored, then restored when the car comes in for clean-up when going ‘off shift’. With enough cameras watching and maybe body cams on service personnel, the employee theft potential could be kept to a minimal level.

  11. Sounds like another iffy business proposition in hopes of sustaining fast growth. Internet satellites, sub-orbital passenger service, and now a self-driving Uber-ish service?

    Taxi service puts a lot of wear and tear on a car – do I want my lovely new Tesla used up by others? Who cleans it? Is the need for frequent cleaning and sometimes repairs included in the Tesla share of income? If someone rents my ‘taxi’ and rides it to the far side of town, or a kid throws up in it, just before I want it to fetch me home from work, do I get free priority use of a Tesla-cab? Must be a bunch more such questions…

    If I were leasing the car I’d worry less about the accelerated wear – but that doesn’t achieve their goal of getting Tesla customers to pay for the cars, and the lease rate would need to be much higher to account for the much faster depreciation.

    They might be better off just making deals with existing taxi companies for a cut of the action, which also avoids creating local opponents who are pals with local politicians. Or if they want to tap into their true-believers, maybe let owners buy stock in a company to finance purchase of the cars – another perq of buying a Tesla.

  12. It’s like those bitcoin mining ASIC’s. If they actually could make money, why would they sell it to you?

    Though in the Tesla case, they can casually work the spectrum from full private ownership for people who can actually afford a car without a lease, to robotaxi PRT via ride hailing app for all the plebs who would been rejected for a car lease (as well as robotaxi rideshare). Tesla owning the whole stack has vertical integration advantages, but that has to be balanced against actual cloud service operators and their own service experience.

    Actually doing the route scheduling for a shared robotaxi service is a Real Problem in computer science, akin to the traveling salesman problem writ large. Some of the early work on this occurred during the last serious push for small single engine jet air taxi services operating from small airports. The patents should be expiring soonish, but then there’s Uber and their operational experience with scheduling (in so much as advertising specific fares to drivers who select a fare out of many, though they are free to do general wide range searches for fares not originally suggested by the system).

    If Uber hadn’t become such a pariah, one could have seen a scenario where they acquired a bankrupt car maker, forcing it to go all electric, and owning the whole stack from a cloud down perspective.

  13. Listening to the challenges that Tesla is taking on, it makes wonder what our highway infrastructure would look like if cameras and computers had come before cars!!

  14. I have talked to Lyft drivers that make around 50k a year. So it is not that unlikely.
    Also Tesla has a user agreement that only allows them to operate a robo taxi fleet based on Teslas.

  15. They could for some limited period. Tesla car output is still very small so people actually can’t buy 10 teslas.

  16. I think Musk should first tackle the problem of what he aptly described as “soul-destroying traffic gridlock“. That might be better addressed not through Tesla-style consumer vehicles, but perhaps rather through those skate-transporters he envisioned with The Boring Company. Imagine if the skates could all coordinate with each other in a congestion situation, to enable faster traffic throughput.

  17. I think Ol’Musky’s timeline is a little optimistic, but if Tesla can get this level of autonomy going in the next few years it will be a major game changer in many ways. Cheap to run self driving vehicles have the potential to cause some significant shifts in how people live & work. For one, self driving taxis should be cheap enough to operate that they will basically kill mass transit in all but the densest locations, like NYC. Already it looks like Lyft & Uber have been cutting into transit use, now imagine that but even cheaper.

    Secondly, longer commutes become viable when your “fuel” cost are 4 cents a mile & you can watch Netflix while you ride. That is going to have all sorts of impact on urban/suburban development. For better or for worse, look for even more urban sprawl.

    Finally, it will probably have an impact on vehicle ownership patterns. I do think that most people will still want to have a car, because of the peak use issue & people want to store stuff in their cars, but it will probably reduce the number of vehicles households have, and may influence them to pick smaller vehicles for day to day use, if renting a bigger vehicle and have it show up at your front door is reasonably easy and cheap for when you do need to do a big hardware store run or the like. On the flipside though, people who don’t buy a big vehicle now because of gas prices might be willing to do so when they are electric; after all, a Model X costs the same to run as a Prius.

  18. This is the same automated fleet plan being worked on by Waymo, GM-Lyft, Ford, and Uber.

    I wish Tesla the best of luck, it is a heavy lift to get to where they want to go.

  19. “By Tesla’s estimates, owners might be able to earn $30,000 in gross revenue from their cars per year. “

    Fun game: find the weasel word!

    (Answer: “might”)

    No. No you cannot ever earn $30,000 in gross revenue per year. Sorry you can’t do that.

    Why? Math. If people start getting near $30k a year in gross revenue then people like me would buy a dozen Teslas, send them on their merry way and flood the robo-taxi market with robo-taxi supply. And when supply skyrockets relative to mostly constant demand prices drop.

Comments are closed.