Meet the self-driving brains working with Ford and Volkswagen

Tech Stack

Bryan Salesky, co-founder and CEO of Argo AI, discusses the intricate hardware and software systems that power autonomous vehicles. Argo AI's tech stack includes a sophisticated mini data center within the car, processing data from various sensors to ensure 360-degree awareness and decision-making capabilities. This setup allows the car to track thousands of objects simultaneously, enhancing safety by predicting future movements of pedestrians and vehicles 1. Salesky emphasizes the importance of encoding local traffic norms into maps, which is crucial for community acceptance of self-driving cars 2.

The advantage with self-driving tech is that our software stack can reason about literally thousands of objects at the same time.

--- Bryan Salesky

This nuanced approach contrasts with the simplistic early visions of self-driving cars, highlighting the complexity and gradual evolution of the technology 3.

   

Data & Mapping

Data and mapping are pivotal in supporting self-driving maneuvers and decision-making processes. Salesky explains that Argo AI utilizes high-resolution maps to help vehicles determine their precise location within lanes, which differs from consumer navigational maps 4. The diversity of data collected from various environments is crucial for refining machine learning algorithms, as not all driving conditions are equal 5.

Data on I-75 is going to be very, very different than data on Broadway.

--- Bryan Salesky

This approach underscores the importance of understanding and predicting the dynamic nature of the driving environment, a concept that has evolved significantly over the years 6.