In the world of autonomous cars, there is a lot of speculation about which companies will come out on top and what factors will contribute to their success. Benedict Evans explores the concept of winner-takes-all effects in this realm and delves into the various layers that play a role in the development and deployment of autonomous vehicles.

Glasp

Hatched by Glasp

Sep 25, 2023

4 min read

0

In the world of autonomous cars, there is a lot of speculation about which companies will come out on top and what factors will contribute to their success. Benedict Evans explores the concept of winner-takes-all effects in this realm and delves into the various layers that play a role in the development and deployment of autonomous vehicles.

One key point that Evans makes is that while the hardware and sensors for autonomy and electric vehicles will likely become commodities, the real differentiating factor lies in the autonomous software that enables a car to navigate without collisions. Additionally, the city-wide optimization and routing systems that allow for the automation of all cars as a system, rather than individually, will be crucial. Finally, the concept of on-demand fleets of "robo-taxis" is also an important aspect to consider.

These three layers - driving, routing & optimization, and on-demand services - are largely independent from a technological standpoint. For instance, one could potentially install the Lyft app in a GM autonomous car and utilize the pre-installed Waymo autonomy module to transport passengers. However, the underlying theme that ties all of these layers together is data.

When it comes to autonomy, there are two types of data that are of utmost importance - maps and driving data. Maps have network effects, meaning that the more autonomous vehicles a company sells, the more frequently and accurately their maps will be updated. This allows for a higher degree of confidence in the vehicle's ability to navigate unknown or changing environments. The second network effect comes from driving data, which is used not only to improve the software's ability to react to different scenarios but also to simulate various situations and test the software's response.

Waymo, for example, benefits from being part of Google, as it reported driving 25,000 real autonomous miles per week and simulated one billion miles in 2016. Tesla, on the other hand, focuses on gathering driving data without the use of LIDAR, which allows them to save time and bypass the need for expensive sensors. However, this approach requires their computer vision software to solve more complex problems.

The winner-takes-all effects in the autonomous car industry are centered around data. Companies that can gather and utilize large amounts of driving data and maintain accurate and up-to-date maps will have a significant advantage over their competitors. This is reminiscent of the PC or Android OEMs, where the tech company creates the network effect through software adoption, leaving the OEMs with near-commodity products.

The question then becomes, how strong is the network effect and how many users or cars are needed to reach a point of diminishing returns? It is likely that Level 5 autonomy, where cars are fully autonomous without any need for manual controls, will emerge gradually from Level 4 autonomy. As the manual controls become less necessary, they will eventually be hidden and removed altogether.

In a scenario where network effects are relatively weak, there may be several companies with viable autonomy platforms. In this case, the car industry would purchase autonomy as a component, similar to how ABS, airbags, or satnav are currently integrated into vehicles.

To navigate the complex landscape of autonomous cars and potentially benefit from winner-takes-all effects, here are three actionable pieces of advice:

  • 1. Focus on data: Invest in gathering and utilizing driving data and maintaining accurate maps. This will give your autonomous car platform a significant advantage over competitors.
  • 2. Embrace network effects: Seek partnerships and collaborations with other companies in the industry to create a robust network effect. This will enhance the value of your product and help drive adoption.
  • 3. Adapt and evolve: Recognize that Level 5 autonomy will likely emerge gradually from Level 4. Embrace this evolution by continuously improving your autonomy platform and reducing the reliance on manual controls.

In conclusion, the winner-takes-all effects in autonomous cars are driven by data. Companies that can gather and utilize driving data effectively, while also maintaining accurate maps, will have a significant advantage over their competitors. By focusing on data, embracing network effects, and adapting to the evolving landscape, companies can position themselves for success in this rapidly developing industry.

Hatch New Ideas with Glasp AI 🐣

Glasp AI allows you to hatch new ideas based on your curated content. Let's curate and create with Glasp AI :)