From Science Fiction to Reality: Tesla’s search for a driverless car
What do you mean your car is driving itself?” said my mom as I showed her a video of how a nearby Tesla came to pick us up completely driverless, through just the press of a button.
But what felt to her like a scene from a science fiction film, was actually the very real consequences of Tesla’s breakthrough and ai-powered driverless system.
Tesla has been able to setup a collective AI network that turns each of their drivers into expert trainers. By simply just driving the vehicle on autopilot mode, each driver reinforces wanted driving behavior from the car and discourages unwanted driving behavior. And because all Teslas are part of a fleet, when one car learns, it passes that knowledge onto t all other Teslas over a neural network, so they all learn together, in what basically constitutes as a hive mind of Teslas.
In what seems like the perfect combination of machine learning and autonomy, the tens of thousands of human drivers can add to a million miles of new data to the network every day. Thereby making what once felt like the pipedream of a truly driverless car into a reality Elon Musk has been quoted as saying is “no more than 10 years away“. In fact in anecdotal reviews, users have already been quoted at document specific instances of the car learning maneuvers it was not previously able to execute (such as taking aggressive turns). It’s definitely learning. It’s safe to say what was once thought to be science fiction is now reality today.
What is most interesting about the potential for this AI-development is that it could have significant effects across, both value creation, value capture and its operating model.
In terms of value creation, a car with significant driverless capability has the ability to complete change and expand the automotive market. From not only being sold to those who currently don’t possess or can’t qualify for a drivers’ license but to even groups of people, who might not currently be able to afford a car on their own, but could choose to pool resources and share a car, without having to worry about driving it between locations (ie. two working professionals who go into the office at different times, could drive the car to the office and then have it drive it back, pick-up the second person and then go to a new destination).
The argument for value capture is even more straightforward. Tesla is now the number one seller of high end premium sedans in North America, at the same time, their superchargers are located across the entire US, allowing for an expanded geographical footprint in their home country. The success is not only local though, in premium markets like Switzerland, the Model S has sold more units than the Mercedes S class, BMW 6 series and Audi A7 combined.
When it comes to the operating model, the impact can be seen throughout every segment of the company. Because the idea of a driverless car creates fear amongst the general population, and generates legal implications in terms of liability for accidents, the Tesla team has had to go even further to build redundancies for every key system to ensure the car continues to drive itself (or alert the driver that they must take control even if one of its systems fails). This means, two centralized computers, two sensor systems, etc. The need for built in redundancies affect the design of the car and therefore its manufacturing processes. It also requires all their engineers to partner with both software and hardware developers to put in place secure systems.
From a sales perspective, it opens up a whole new realm for customer acquisition. Since Tesla currently owns all of its dealerships, it focuses on a B2C model. However a driverless would eliminate the need for individuals to be drivers and would be much more appealing for businesses. Early signs of this can already be seen in Europe, where in Madrid, Teslas make up over 60% of existing Ubers. A move that goes even deeper into driverless technologies could attract cab companies, and public transportation agencies into adopting Teslas, forever changing it into a B2B company.