3 ms·

One thing I'll be curious about here over the long run is what winds up being the actual effective cost per mile of a taxi-style ride - my understanding is coun

One thing I'll be curious about here over the long run is what winds up being the actual effective cost per mile of a taxi-style ride - my understanding is counting deprecation and repair costs, Uber's pricing has already been at or below cost, and most drivers are effectively underwater when you price in the cost of wear and tear on the vehicle itself. Custom-built ultra-high-duty-cycle vehicles a la the mail trucks and the London cabs should help here, but I think it's an open question just how cheap you can really make a ride, and I'm not sure the meat in the drivers seat is the primary driver of cost at this point.

bigyikes 2y ago
Vehicle depreciation doesn’t matter unless you’re selling the thing. Wear and tear will be reduced with an all-electric drivetrain, which is uncommon for Ubers outside of California. Costs will be further reduced by bulk maintenance and solar charging. Autonomous vehicles will also have extremely high utilization rates since they can run 24 hours a day, versus a cab or Uber vehicle which sits idle, losing money.
- com2kid 2y ago
  > Autonomous vehicles will also have extremely high utilization rates since they can run 24 hours a day, versus a cab or Uber vehicle which sits idle, losing money. Math doesn't work out there. You need N number of vehicles to give good service (low wait times) during peak hours and special events, and a much smaller number of vehicles the rest of the time. Only a small portion of a fleet will need to run 24/7. Now that said, they can cycle vehicles in and out of 24 hour shifts to spread the wear and tear more evenly across the fleet, assuming there is a benefit to doing that.
  - ra7 2y ago
    The vehicles can also be redeployed during off-peak hours for delivery services. They’ve previously also talked about private subscriptions for individuals.
- roughly 2y ago
  I expect some tuning around torque curves and the like, but electric drivetrains aren’t a pure positive for wear and tear - the heavier weight plus torque and general drivetrain characteristics are murder on tires, suspension, and several other wear items on the vehicle. High duty, and especially high duty from taxi-style driving, is likely to impact wear as well. You’re right about the benefits of bulk maintenance, although I’m skeptical that a taxi operator will be able to pull off pure solar charging - typical electric car batteries are on the order of 40-60kWh, which is about what an average home uses in a day or two. Scale that up to a reasonably-sized fleet of electric vehicles, and I suspect any fleet operator will be buying their juice. The utilization rate on the vehicles will be interesting - browsing around a bit looks like an average cab runs between 70-100k miles per year (the car, not the driver - most taxis run multiple shifts), so call it ~200-300mi per day, so they ought to be able to make it the whole day on a single charge. Once you own the vehicle, though, it’s pay per use - as you note, if you’re not planning to sell it, it’s not really losing money sitting around. It’s not making money, but it’s also not incurring costs from operation.
- JumpCrisscross 2y ago
  > Vehicle depreciation doesn’t matter unless you’re selling the thing Depreciation is replacement cost smoothed out. If a business is unprofitable after depreciation it cannot organically replace its capital equipment. That said, I’m sceptical of the Uber and Lyft being unprofitable for drivers after depreciation. Doubly so in the cities Waymo targets.
- RivieraKid 2y ago
  Utilization will be higher but possibly not by much because demand is not constant but peaks when people move to or from work. I wonder how does the higher utilization affect costs. For example, if you have a car that can last 400k miles, how much do you save if you finance it for 2 years vs for 6 years? One thing to consider is that there's miles-driven depreciation and time depreciation. Time depreciation means that the vehicle loses value over time even if you don't use it. A brand new Waymo Chrysler Pacifica will have much less value in 20 years than today.
jfengel 2y ago
I wonder if you can save significant money on the vehicle itself. You could even design a one seat version, which I suspect would be the majority of rides. It would also save some gas. I know it wouldn't be enormous, and probably zero savings on maintenance. But perhaps you could save 25 percent all told?
- RivieraKid 2y ago
  I would much prefer if the trend was multiple people sharing a vehicle because less traffic would make cities nicer places to live. This technology can shrink the comfort gap between public transport and using a taxi.
  - Zigurd 2y ago
    It's a good bet that vehicle form factors will proliferate. We are at the point of having the first purpose-built vehicles. When there are millions of them they will probably vary in size and capacity. But only if smaller ones can be cheaper and usage patterns support high capacity vehicles, which probably won't be uniformly true across different cities.
ra7 2y ago
One of Waymo’s co-CEOs went over the estimates in a recent podcast: Vehicle lifetime: 400,000 miles Upper bound vehicle cost: $100,000 (though I expect Zeekr to manufacture it for way less) Cost: $0.25 per mile There’s room for margin compared to Uber since they can take advantage of economies of scale (high utilization, centralized operations/maintenance and so on).
- roughly 2y ago
  If that’s just lifetime/cost, that’s not covering energy or repairs.
- RivieraKid 2y ago
  I tried to calculate this a few months ago and the result was $0.7 with room for improvement over time. I had $0.25/mi for vehicles, 0.14 for labor costs (remote and road-side assistance, charging, cleaning, management, etc.), 0.05 electricity, 0.05 for buildings and parking lots, 0.025 vehicle maintenance and insurance. That's 0.52 in total. Assuming 75% miles are paid, the cost per paid mile is $0.7. I can imagine significant economies of scale, e.g. 20% cost reduction for each doubling of deployment, so maybe $0.1 after a few decades?

bigyikes 2y ago

Vehicle depreciation doesn’t matter unless you’re selling the thing. Wear and tear will be reduced with an all-electric drivetrain, which is uncommon for Ubers outside of California. Costs will be further reduced by bulk maintenance and solar charging. Autonomous vehicles will also have extremely high utilization rates since they can run 24 hours a day, versus a cab or Uber vehicle which sits idle, losing money.

com2kid 2y ago
> Autonomous vehicles will also have extremely high utilization rates since they can run 24 hours a day, versus a cab or Uber vehicle which sits idle, losing money. Math doesn't work out there. You need N number of vehicles to give good service (low wait times) during peak hours and special events, and a much smaller number of vehicles the rest of the time. Only a small portion of a fleet will need to run 24/7. Now that said, they can cycle vehicles in and out of 24 hour shifts to spread the wear and tear more evenly across the fleet, assuming there is a benefit to doing that.
- ra7 2y ago
  The vehicles can also be redeployed during off-peak hours for delivery services. They’ve previously also talked about private subscriptions for individuals.
roughly 2y ago
I expect some tuning around torque curves and the like, but electric drivetrains aren’t a pure positive for wear and tear - the heavier weight plus torque and general drivetrain characteristics are murder on tires, suspension, and several other wear items on the vehicle. High duty, and especially high duty from taxi-style driving, is likely to impact wear as well. You’re right about the benefits of bulk maintenance, although I’m skeptical that a taxi operator will be able to pull off pure solar charging - typical electric car batteries are on the order of 40-60kWh, which is about what an average home uses in a day or two. Scale that up to a reasonably-sized fleet of electric vehicles, and I suspect any fleet operator will be buying their juice. The utilization rate on the vehicles will be interesting - browsing around a bit looks like an average cab runs between 70-100k miles per year (the car, not the driver - most taxis run multiple shifts), so call it ~200-300mi per day, so they ought to be able to make it the whole day on a single charge. Once you own the vehicle, though, it’s pay per use - as you note, if you’re not planning to sell it, it’s not really losing money sitting around. It’s not making money, but it’s also not incurring costs from operation.
JumpCrisscross 2y ago
> Vehicle depreciation doesn’t matter unless you’re selling the thing Depreciation is replacement cost smoothed out. If a business is unprofitable after depreciation it cannot organically replace its capital equipment. That said, I’m sceptical of the Uber and Lyft being unprofitable for drivers after depreciation. Doubly so in the cities Waymo targets.
RivieraKid 2y ago
Utilization will be higher but possibly not by much because demand is not constant but peaks when people move to or from work. I wonder how does the higher utilization affect costs. For example, if you have a car that can last 400k miles, how much do you save if you finance it for 2 years vs for 6 years? One thing to consider is that there's miles-driven depreciation and time depreciation. Time depreciation means that the vehicle loses value over time even if you don't use it. A brand new Waymo Chrysler Pacifica will have much less value in 20 years than today.

jfengel 2y ago

I wonder if you can save significant money on the vehicle itself. You could even design a one seat version, which I suspect would be the majority of rides. It would also save some gas. I know it wouldn't be enormous, and probably zero savings on maintenance. But perhaps you could save 25 percent all told?

RivieraKid 2y ago
I would much prefer if the trend was multiple people sharing a vehicle because less traffic would make cities nicer places to live. This technology can shrink the comfort gap between public transport and using a taxi.
- Zigurd 2y ago
  It's a good bet that vehicle form factors will proliferate. We are at the point of having the first purpose-built vehicles. When there are millions of them they will probably vary in size and capacity. But only if smaller ones can be cheaper and usage patterns support high capacity vehicles, which probably won't be uniformly true across different cities.

ra7 2y ago

One of Waymo’s co-CEOs went over the estimates in a recent podcast: Vehicle lifetime: 400,000 miles Upper bound vehicle cost: $100,000 (though I expect Zeekr to manufacture it for way less) Cost: $0.25 per mile There’s room for margin compared to Uber since they can take advantage of economies of scale (high utilization, centralized operations/maintenance and so on).

roughly 2y ago
If that’s just lifetime/cost, that’s not covering energy or repairs.
RivieraKid 2y ago
I tried to calculate this a few months ago and the result was $0.7 with room for improvement over time. I had $0.25/mi for vehicles, 0.14 for labor costs (remote and road-side assistance, charging, cleaning, management, etc.), 0.05 electricity, 0.05 for buildings and parking lots, 0.025 vehicle maintenance and insurance. That's 0.52 in total. Assuming 75% miles are paid, the cost per paid mile is $0.7. I can imagine significant economies of scale, e.g. 20% cost reduction for each doubling of deployment, so maybe $0.1 after a few decades?