“One thing is that we now prepare it at home, so instead of relying on long deadline and long -process suppliers and sending back requirements, we manufacture software,” said ADAS developer, Zanet’s VP, Eric Query, owned by Volvo.
“If there is anything, we solve it in a day,” said Qungh. “This is very fast. So we repeat very fast. As Beaches said, we are all testing with new software every day. So the speed of innovation is mainly different than ever before, and the way we try to use it really is to promote safety and never know the rate of crash.”
One of the benefits of moving to SDV is that it is very easy to imitate them because the entire software stack can be practically operated. This is why Volvo has made one of Europe’s largest data centers: so that they can be able to run these Sims.
Like many other companies there, Volvo has turned to AI to accelerate the development process. But how does a company aware of the safety like Volvo know that it can rely on the production of algorithms at the end?
“Gaosi spilting is a technology where we can take a point, a traffic scenario, and it can explode in thousands of or tens of thousands of landscapes from real -world data,” said Quengh. “And then we can add a scene to a thousand different scenarios, and then we can connect the imitation and test our software against it.”
Developers of sovereign vehicles have long been imitating an unrealistic engine environment. “It’s very visual. It works for camera data. But here we are investigating leather data, camera data, radar data, and we create this scene with the nerve net, and then manipulate and use closed loop fake,” “So this is such a way, be able to test your software in a very fast, large, large quantity of different scenarios that are representatives of the real world.”