Audi Forms Partnership with Israeli Startup Cognata

author: Michael Cheng   

Ensuring autonomous vehicles are fit for public roads is a painstaking task that requires a lot of patience. To speed up this aspect of development, many companies in the automotive space have turned to virtual testing environments, such as the simulations offered by Israeli startup Cognata.

The company recently announced a partnership with Audi to develop autonomous cars. Part of the collaboration entails the integration of Cognata's virtual platform with Audi's Autonomous Intelligent Driving (AID) program.

Autonomous Intelligent Driving (AID) Program

For Audi, the partnership may help the automaker bring driverless features to its existing lineup of cars faster and at lower costs. Cognata offers a wide range of virtual cities that closely resemble real locations, including San Francisco and Munich, which allows Audi to focus on vehicle deployment over the details and mechanisms of the simulation.

As mentioned earlier, the startup's products and expertise will be funneled through the car manufacturer's AID program. Implementation for the program includes all stages of development, as well as large-scale vehicle production.

"At AID, we are convinced that simulation is a key tool to increase our development speed and a necessary one for the validation of our product and for proving it is safe," explained AID Chief Technology Officer Alex Haag.

"After exploring various solutions, we decided that partnering with Cognata is the fastest way to reach these goals."

Automotive Testing in Virtual Environments

The startup's virtual proving grounds makes use of artificial intelligence and deep learning for realistic elements inside the simulation. Powered by the cloud, developers can run new driving maneuvers over and over again without needing to manually reset the environment after each trial. As a result, the flow of testing is streamlined and less time is spent on setup.

Cognata's deep-learning variables consist of three different layers. Two of the three layers are designed to recreate elements and objects in the vehicle's surroundings, such as lane markings and trees. The third layer provides sensing capabilities on the autonomous car (in a virtual environment). Interestingly, developers have already modeled 40 driverless sensors for simulations, which makes the entire experience very realistic.

Moreover, the environment is displayed in 3D and enables developers to speed up testing via a fast-forward option. Outdoor environments can be configured seamlessly to mimic slippery roads, low-visibility conditions and more. For localized simulations, the driving behavior or other cars on the road may also be customized to suit the requirements of the trial.

"Simulation technology can help greatly reduce autonomous vehicle testing costs for carmakers," said Cognata Chief Executive Officer Danny Atsmon.

"If you look at other startups in the simulation world, they're recreating very small scenes, nothing close to what we have here."

According to Atsmon, without virtual testing platforms for autonomous cars, it would take roughly 600 years to train driverless vehicles with competency that matches a human driver.

Cognata is backed by several key players in auto and aviation. Airbus Ventures, Emerge and Maniv Mobility (supported by Jaguar Land Rover and Valeo) are currently invested in the startup.

Michael Cheng

Michael Cheng is a legal editor and technical writer with publications for Blackberry ISHN Magazine Houzz and Payment Week. He specializes in technology business and digesting hard data. Outside of work Michael likes to train for marathons spend time with his daughter and explore new places.