Hello fellow TMC users!
I was a bit surprised to not see an analysis of profit potential that Elon quoted at $30k USD/yr. As such, I made my own, and I welcome any feedback or criticism. Any is welcome!
Docs link: Tesla Autonomy April 2019
For a version without images, for those who do not wish to load a google docs, I have copy-pasted the body here:
On April 22nd, Tesla made this incredible presentation about full self driving (FSD) on Tesla Autonomy Day. It was followed by some pretty advanced videos.
Autonomy Full Presentation
FSD Demo Video
Although I have long been under the belief that autonomous ride hailing is the future, I felt compelled to write about it again, with math, difficulties, and extra new information.
No other production car on the road today has the hardware capable of FSD, which is why it is a financially foolish idea to buy any vehicle other than a Tesla today.
Even if the timeline for unsupervised FSD could be off by years, the hardware is there, and thanks to over the air updates, the software will arrive when it is ready.
Originally - and naively - I thought that I would be able to purchase new Tesla’s for my autonomous fleet after my current Model 3 started making money. This was faulty logic for a number of reasons. As Elon Musk pointed out on Twitter - Elon Musk on Twitter - , the software for FSD will become substantially more expensive over time, as it gets closer to validation and to being a revenue source for owners. Given the massive potential revenue source, I can easily imagine that the FSD software will eventually cost $100,000 or more, per vehicle. It does not make sense for Tesla to sell their vehicles - including the software - for as cheap as they do today, when they have the possibility of making hundreds of thousands of dollars per vehicle by keeping it to themselves. This is why buying it now makes the most sense, but requires the most risk, and financial backing.
Let’s delve into what kind of income a Robotaxi could earn now.
Currently, on the Canadian Tesla store, you can buy a standard plus with FSD for $60,300. Add the dest/doc fee and taxes, and you’re looking at $69,608 for a brand new standard plus Model 3 with FSD package. If you were looking to do a $7500 down payment, and financed for 8 years at 4.4%, you’re looking at a monthly payment of $768.66. That is already competitive with most new ICE cars when you consider gas savings over a long time, but we’re here to talk about the network, and how even if that may be a bit more expensive than that new Toyota Camry or Honda Civic you were looking at, it’s an investment into the future instead of a depreciating asset.
Based on numbers I researched a few years ago and tweaking for today’s rates, along with numbers given by Tesla, I came up with a few data points, which are labeled neatly in the graph below. Vehicle cost and monthly payments are listed above.
For income:
(Unfortunately, I cannot figure out how to make a graph in the forum post. Please follow the following link to see the graph in graph form. Otherwise it will be formatted as best I can for this format)
Sheets link: Tesla Spreadsheet
Constant variables/assumptions:
Costs:
Tesla vehicle cost - $69,608
Monthy car payment on $62108 loan with 4.4% interest over 8 years - $769
km/hour averaged for 16 hrs/day - 26
Electricity costs (kWh) - $0.13
Efficiency, counting energy loss (wh/km) - 180
Days a year not driving (Holidays, shop days, etc) - 20
Tire cost per 60,000 km - $1,000
Maintenance cost per 20,000 km - $500
Maintenance cost per km - 0.03
Income:
Uber costs ~$2.50 per km of ride-share
Price per km if Tesla charged less than half - $1.00
km/shift for 16 hrs/day - 416
Wasted % - 50%
km/shift of income - 208
Tesla's share - 30%
Time per day Tesla Share - 16 hrs per day
Costs:
km/year - 143500
kWh/year - 25,830
cost for kWh/year - 3,357.90
Maintenance per year - 3,587.50
Tire annual costs - 2391.67
Insurance - 2500
Annual car payments - $9,224
Total annual cost - 21,060.99
Income:
Per shift - 208
Per year - 71760
After dispatch share - 50232
Net yearly income - 29,171.01
As you can see from the above, and from the presentation, there will be significant amount of money to be made from owning a Tesla with FSD equipped. There are, of course, certain roadblocks including time it would take for full legislation.
As such, this makes sense for anyone who may have the financial means to use a Tesla as their primary vehicle until then, but it may be a bit of a money sink until then.
Or is it?
With the use of rental car companies, or sites such as Turo, one is able to monetize their Tesla prior to FSD, though not to the same extent. If no deal could be acquired with a rental service, then Turo could help to break even - or even gain a small profit - during the process of waiting for your investment - the FSD Tesla - to mature. I, unfortunately, cannot give numbers even close to accurate for this as I simply do not have the dataset to rely on.
Ontario
Surprisingly, Ontario actually already allows level 3 autonomous vehicles on the roads for testing purposes. There is a pilot permit application form to go through for higher levels of autonomy, and allows - under certain conditions - higher level autonomy to be used with remote operators. For more information, please visit: Automated Vehicles – Driving Innovation in Ontario
Adding Autonomy Later:
As mentioned earlier, FSD software upgrade will get significantly more expensive over time. If that software were to go up to $100,000 or more, who would ever be able to front that much money, especially if they have no guarantee that they will be able to recoup the cost before Tesla changes their mind and does not allow their “older” (at that point) vehicle to be part of their fleet?
This is why I propose an alternate method of payment to Tesla. Using above examples, and a cost of $100,000 for FSD, I would hope that Tesla allows owners who have not yet purchased FSD to still allow their vehicles onto the fleet, albeit at a significantly reduced rate of earnings. The above example mentioned 30% Tesla cut. To incentivize non-FSD buyers to use the network, Tesla could take a 80% cut, still leaving 20% to the person who does not yet have FSD software. That would continue until Tesla has earned their $100,000. In the above example, that would be close to two years, but the owner would still be receiving a discount on their monthly payment, paying a total of $6,708 yearly instead of over $12,000.
I believe that would still make sense and entice owners to use the fleet service even without owning FSD, given that they would still get a future profit potential.
Risks, Roadblocks:
Of course, there are risks to the above.
In terms of FSD itself, some possible issues include:
I was a bit surprised to not see an analysis of profit potential that Elon quoted at $30k USD/yr. As such, I made my own, and I welcome any feedback or criticism. Any is welcome!
Docs link: Tesla Autonomy April 2019
For a version without images, for those who do not wish to load a google docs, I have copy-pasted the body here:
On April 22nd, Tesla made this incredible presentation about full self driving (FSD) on Tesla Autonomy Day. It was followed by some pretty advanced videos.
Autonomy Full Presentation
FSD Demo Video
Although I have long been under the belief that autonomous ride hailing is the future, I felt compelled to write about it again, with math, difficulties, and extra new information.
No other production car on the road today has the hardware capable of FSD, which is why it is a financially foolish idea to buy any vehicle other than a Tesla today.
Even if the timeline for unsupervised FSD could be off by years, the hardware is there, and thanks to over the air updates, the software will arrive when it is ready.
Originally - and naively - I thought that I would be able to purchase new Tesla’s for my autonomous fleet after my current Model 3 started making money. This was faulty logic for a number of reasons. As Elon Musk pointed out on Twitter - Elon Musk on Twitter - , the software for FSD will become substantially more expensive over time, as it gets closer to validation and to being a revenue source for owners. Given the massive potential revenue source, I can easily imagine that the FSD software will eventually cost $100,000 or more, per vehicle. It does not make sense for Tesla to sell their vehicles - including the software - for as cheap as they do today, when they have the possibility of making hundreds of thousands of dollars per vehicle by keeping it to themselves. This is why buying it now makes the most sense, but requires the most risk, and financial backing.
Let’s delve into what kind of income a Robotaxi could earn now.
Currently, on the Canadian Tesla store, you can buy a standard plus with FSD for $60,300. Add the dest/doc fee and taxes, and you’re looking at $69,608 for a brand new standard plus Model 3 with FSD package. If you were looking to do a $7500 down payment, and financed for 8 years at 4.4%, you’re looking at a monthly payment of $768.66. That is already competitive with most new ICE cars when you consider gas savings over a long time, but we’re here to talk about the network, and how even if that may be a bit more expensive than that new Toyota Camry or Honda Civic you were looking at, it’s an investment into the future instead of a depreciating asset.
Based on numbers I researched a few years ago and tweaking for today’s rates, along with numbers given by Tesla, I came up with a few data points, which are labeled neatly in the graph below. Vehicle cost and monthly payments are listed above.
- Elon mentioned a maximum 16 hours a day of available time per day of ride-share, counting for lower demand times, charging, etc.
- He also estimated 16 mph (26 km/h) as the average speed when counting for stops, pickups and dropoffs, etc.
- Where I live, in Ottawa, my electricity prices average to $0.13 per kwh
- Although Tesla Model 3’s efficiency is rated at 150 wh/km, I raised that to 180 wh/km to try to account for time of year average, efficiency loss charging, etc. It may still be a bit inaccurate
- Accounted for some days that just couldn’t be driven during the year, such as severe weather that FSD does not know how to do yet, or repairs for whatever reason
- Accounted for new tires every 60,000 km
- Accounted for a “just in case” maintenance cost. Tesla maintenance costs are extremely low, but I wanted a buffer either way.
For income:
- Uber costs “$2-3 per mile”, extrapolated to CAD and km was around $2.50 per km
- Assumed that Tesla network would charge less than half
- Used the above to estimate mileage put on car, amount of wasted mileage getting to destinations, and then accounted for an approximately 30% Tesla share off the top.
- I then changed everything to yearly to come up with a gross income, expenses, and net income.
(Unfortunately, I cannot figure out how to make a graph in the forum post. Please follow the following link to see the graph in graph form. Otherwise it will be formatted as best I can for this format)
Sheets link: Tesla Spreadsheet
Constant variables/assumptions:
Costs:
Tesla vehicle cost - $69,608
Monthy car payment on $62108 loan with 4.4% interest over 8 years - $769
km/hour averaged for 16 hrs/day - 26
Electricity costs (kWh) - $0.13
Efficiency, counting energy loss (wh/km) - 180
Days a year not driving (Holidays, shop days, etc) - 20
Tire cost per 60,000 km - $1,000
Maintenance cost per 20,000 km - $500
Maintenance cost per km - 0.03
Income:
Uber costs ~$2.50 per km of ride-share
Price per km if Tesla charged less than half - $1.00
km/shift for 16 hrs/day - 416
Wasted % - 50%
km/shift of income - 208
Tesla's share - 30%
Time per day Tesla Share - 16 hrs per day
Costs:
km/year - 143500
kWh/year - 25,830
cost for kWh/year - 3,357.90
Maintenance per year - 3,587.50
Tire annual costs - 2391.67
Insurance - 2500
Annual car payments - $9,224
Total annual cost - 21,060.99
Income:
Per shift - 208
Per year - 71760
After dispatch share - 50232
Net yearly income - 29,171.01
As you can see from the above, and from the presentation, there will be significant amount of money to be made from owning a Tesla with FSD equipped. There are, of course, certain roadblocks including time it would take for full legislation.
As such, this makes sense for anyone who may have the financial means to use a Tesla as their primary vehicle until then, but it may be a bit of a money sink until then.
Or is it?
With the use of rental car companies, or sites such as Turo, one is able to monetize their Tesla prior to FSD, though not to the same extent. If no deal could be acquired with a rental service, then Turo could help to break even - or even gain a small profit - during the process of waiting for your investment - the FSD Tesla - to mature. I, unfortunately, cannot give numbers even close to accurate for this as I simply do not have the dataset to rely on.
Ontario
Surprisingly, Ontario actually already allows level 3 autonomous vehicles on the roads for testing purposes. There is a pilot permit application form to go through for higher levels of autonomy, and allows - under certain conditions - higher level autonomy to be used with remote operators. For more information, please visit: Automated Vehicles – Driving Innovation in Ontario
Adding Autonomy Later:
As mentioned earlier, FSD software upgrade will get significantly more expensive over time. If that software were to go up to $100,000 or more, who would ever be able to front that much money, especially if they have no guarantee that they will be able to recoup the cost before Tesla changes their mind and does not allow their “older” (at that point) vehicle to be part of their fleet?
This is why I propose an alternate method of payment to Tesla. Using above examples, and a cost of $100,000 for FSD, I would hope that Tesla allows owners who have not yet purchased FSD to still allow their vehicles onto the fleet, albeit at a significantly reduced rate of earnings. The above example mentioned 30% Tesla cut. To incentivize non-FSD buyers to use the network, Tesla could take a 80% cut, still leaving 20% to the person who does not yet have FSD software. That would continue until Tesla has earned their $100,000. In the above example, that would be close to two years, but the owner would still be receiving a discount on their monthly payment, paying a total of $6,708 yearly instead of over $12,000.
I believe that would still make sense and entice owners to use the fleet service even without owning FSD, given that they would still get a future profit potential.
Risks, Roadblocks:
Of course, there are risks to the above.
In terms of FSD itself, some possible issues include:
- Legislation taking forever to get passed despite overwhelming data. Governments work slowly.
- Knee-jerk reactions instead of looking at data, when issues DO arise. If someone were to die in an autonomous vehicle, even if data proves that they were 100 times more likely to die by a human driver, legislation against automation may go against logic and reason.
- Major lobbying against automation from other car manufacturers who are not yet caught up, or taxi/ride hailing groups, or fossil fuel industry, or other industries that this would negatively impact
- Someone accidentally spills their coffee in your vehicle. It is now out of service to be cleaned. Need cleaning services at the ready.
- Same example as above but with vomit or other bodily fluids.
- Even with interior camera, these will still happen
- In Ottawa, with taxi strikes against Uber, the taxi drivers instigated violence, and even smashed Uber car windows - with occupants inside. There is no telling what they would do to driverless vehicles, even with cameras recording their every move.
