- Volvo’s partnership with nvidia goes beyond chips in cars
- Gaussian Splatting creates 3D scenes with high faith
- The technique can recreate “edge cases” to train models faster
Volvo uses a new AI technique called ‘Gaussian Splatting’ to train his vehicles and speed up its target of zero collisions on the roads – and that’s all thanks to its newly expanded partnership with Nvidia.
Last month, we reported that the upcoming Volvo ES90 will be the most powerful car it has ever created in terms of core computing capacity due to the one that packs a double NVIDIA AGX ORIN configuration.
Now the company has revealed how this kind of supercomputing also helps it to train its advanced driver aid systems faster.
Volvo claims that it can now synthesize incidental data collected by the advanced sensors in its latest vehicles, such as emergency braking, sharp control or manual intervention.
This then allows the company to reconstruct and explore them in new ways to better understand how events can be avoided.
The new method is called Gaussian-Platting, and it allows the company’s software to produce realistic 3D scenes with high faith and topics from the real world.
Once these scenes are created, Volvo’s engineers can manipulate them to generate a number results. The examples of the video clip provided by the Swedish marking are freakishly realistic.
It corresponds to a human learning how to skateboard by playing Tony Hawks Pro Skater For hours and hours on the end.
“We can choose one of the rare edge cases and explode it in thousands of new variations of the scenario to train and validate our models against,” explains Alwin Bakkenes, head of Global Software Technique at Volvo Cars.
Bakkenes says this has the potential to unlock a scale that Volvo has never had before and even to catch edge cases before they happen in the real world.
Now train computers computers
Gaussian splash is a relatively new 3D reproduction technique that is not dependent on neural networks, as opposed to more complex methods such as neural radiation field (NERFs).
This makes it possible to create incredibly complex 3D scenes in real time. The technique is currently being explored in several industries, from games to interactive app development.
Volvo’s use of advanced Lidar, Sensor and High-Definition Camerate Technology, first displayed in the EX90, collects data that can then be reproduced in a manipulated 3D model that allows its engineers to then train the vehicle’s AI to work better in the real world.
There was some disappointment when the EX90 was launched when its Lidar technology would remain offline for consumer use effectively for just collecting data until Volvos Compute Power was at a level where the company was happy to introduce ADAs systems that depended on the sensor suite and software stack.
Fortunately, the newly announced partnership with NVIDIA will help the Swedish marking that is synonymous with road safety in realizing its vision of zero collisions and driver assistant systems that actually help, rather than just fooling.
In addition, the company has also stated that Early Ex90 models will be updated with the double NVIDIA AGX ORIN system on a chip setup so that they can also make the most of the latest developments in autonomous driving and ADAS systems.
