How Will Tesla’s Robotaxi Network Work?

During Tesla’s autonomy day for investors on the 22nd April, Musk announced that he was confident there would be a network of Robotaxis up and running by the end of 2020. That sounds great, but investors are rightfully sceptical, resulting in a drop in TSLA stock.

Elon Musk does have a tendency to set unlikely deadlines for almost all of his projects, such as the Model 3 production ramp, or the SpaceX Falcon 1 launch. However, there is reasoning behind this. First, it encourages the team to work harder to get unbelievable amounts done in a short time. And second, it is likely an attempt to attract investors.

So, is this just Elon being Elon or is it actually something we could be seeing in the future?

How does a Robotaxi work?

Naturally, there are many competitors, including Waymo, googles self-driving car project. Furthermore, pretty much every major automobile manufacturer right now is developing their own self-driving tech, including BMW and Volkswagen. Of course, they won’t all go down the Robotaxi route, but Tesla still has competition.

To reach a Robotaxi level of autonomy, cars are going to need to be at autonomy level 5. That’s the top of a series of 5 levels set by the NHTSA in the USA. Here are the different levels:

  • Level 0 – No autonomy.
  • Level 1 – Driver assistance, such as adaptive cruise control.
  • Level 2 – Functions such as automated acceleration and steering, but the driver must remain alert at all times. (Where many autonomous systems are now)
  • Level 3 – A driver is not necessary, but must be able to take control of the vehicle at all times. (Where Tesla is).
  • Level 4 – Capable of performing all driving functions under certain conditions. (That’s where Waymo operates).
  • Level 5 – All driver functions in all conditions, no driver needed.
An example of a self driving car interior. Personal space.
  • Save
An example of a self driving car interior

Hopefully, customers in Robotaxis will have their own space, free from distractions and safety drivers. As a result, time spent commuting could be enjoyable, doing work, watching TV, or relaxing. Add to that advances in comfort and back out windows and you have yourself your own private space.

Lidar vs vision?

Lidar is a surveying method which measures precise distances by illuminating a target with pulsed laser lights and measuring the return time and wavelength. For example, an object which is further away will have a larger return time

Waymo's self driving car project with lidar on top.
  • Save
Google’s Waymo with the lidar dome on top.

That sounds great, and it is for mapping the environment. However, it comes with some major drawbacks.

Firstly, it costs colossal amounts of money. This is okay, but taxi ride prices are going to increase in order to pay that back. Secondly, it’s quite large and looks unattractive. Contrast that to a Tesla, where you wouldn’t know that there were 8 cameras hidden around the body and the appearance battle seems to be lost already.

However, appearance doesn’t really matter so let’s compare lidar to the vision system which Tesla believes is the future.

Tesla’s vision system uses a series of 8 cameras and 12 ultrasonic sensors to determine the shape and scale of the environment. This is relatively cheap compared to lidar and is much more discrete.

Additionally, it can be used in conjunction with AI for advanced pathfinding and environment recognition, making the system act more like a human driver. AI means that the car can adapt to natural roads, whereas with lidar, the roads need to adapt partially to the car.

One disadvantage with vision systems is the amount of processing power needed to run them. To solve that problem, Tesla developed a fully redundant self-driving computer which is capable of processing 2.5 billion pixels per second. That’s an enormous amount of processing power which can handle high-resolution video feedback and decision making from 8 cameras at once.

Is the Robotaxi System Financially Viable?

The cost of running a Robotaxi, demonstrated by Musk.
  • Save
Costs of running the Robotaxi network, as presented by Musk.

From my initial observations, I think that Tesla’s business model here is far superior to Waymo’s. Here’s why:

Currently, Tesla cars have logged data for over 1 billion miles on autopilot, feeding the AI to generate new solutions. Waymo has a few 10s of cars out in a few cities across the US. Although they have logged quite a few miles, Tesla is far ahead here.

Next, Tesla is positioning their service as a way of the customer making money, not Tesla directly. Cleverly, Tesla still makes money from the sales of the cars and will take a 25-30% cut of the taxi fare. Whereas Waymo has poured huge amounts of money into the cars and their development, meaning the cost of using them is likely to be higher than the cost of using a Tesla.

In the autonomy event, Tesla said that its ride-sharing service would cost about $0.18 per mile in operating costs. That’s cheap. Of course, they will add a profit margin on to that, but I would imagine it will remain cheaper than Waymo.

In summary, I think Tesla’s business model is superior because it is creating profitability for itself by creating success for its customers.

How would a Robotaxi work?

During the Tesla event, the team outlined that we only used our cars for a few hours a day. All of the rest of the time, your car could be earning you money with the Robotaxi network. The first stage is to allow your car to accept ride sharing customers from the Tesla app.

Next, the customer(s) would go on the Tesla app (just like the Uber or Lyft apps) and request to go to a location. The car would autonomously drive there and pick up the customer(s).

Robotaxi ride sharing app demonstration
  • Save
The Tesla Ride Sharing App of the Future.

After that, the car would drive to the location by itself and drop the customer(s) off. I’m presuming that when the car needs to charge, it will stop ride sharing. Obviously, the car cannot go to a supercharger and plug itself in!

Hopefully, the owner would be able to set a minimum range limiter, so the car would stop ride sharing at say 40 miles left (you don’t want your car to return with the battery dead)!

Finally, if there aren’t enough people sharing their cars in certain areas, Tesla would deploy it’s own dedicated vehicles to act as a replacement.

Conclusion – Should shareholders be worried?

So is Elon just being Elon again? Yes and no. His deadlines are usually unrealistic, but they do push his team to work harder and he achieves them in the end. An example of that is the solar roof, which is now coming into mass production.

Many sceptics say that self-driving technology is years away, but I wouldn’t put it past Elon. Many people though reusable rockets were not possible, and we are now seeing all 3 of the Falcon Heavy boosters come and land.

Where this may be a slightly optimistic deadline, I think that Tesla will still pull it off, creating a very profitable business. After all, shareholders should be in for the long term ethos, not the current mess of financials.

Tesla’s vision is to accelerate the advent of sustainable transport by bringing compelling mass market electric cars to market as soon as possible. If that’s what they are focusing on, shareholders shouldn’t be greatly worried.

Moreover, Tesla was aiming for profitability later this year, not in the first quarter. If Tesla can make it through the financial problems they are facing now, we could be seeing Robotaxis soon and it would be revolutionary.

Share via
Copy link
Powered by Social Snap