Wayve CEO shares its key ingredients for scaling autonomous driving technician

Wayve co -founder and CEO Alex Kendall sees the promise to market the technology of his autonomous start -up car. That is, if Wayve adheres to his strategy to ensure that his automated driving software is cheap to implement, hardware agnostic and can be applied to advanced driver, robotaxis and even robotics systems.

The strategy that Kendall has exposed during NVIDIA GTC ConferenceIt starts with a training approach managed from end to end. This means that what the system “sees” through different sensors (such as cameras) directly becomes the way it moves (such as deciding to stop or turn left). In addition, this means that the system does not need to rely on HD cards or a software-based rules, as there are earlier versions of AV Tech.

The approach attracted investors. Wayve, which started in 2017 and has raised more than $ 1.3 billion In the last two years he has planned to license his self -control software for automotive and naval partners, such as UberS

The company has not yet announced car partnerships, but a spokesman told TechCrunch that Wayve is in “Strong Discussions” with multiple OEM manufacturers to integrate its software into a number of different types of vehicles.

His cheap software is crucial to achieving these transactions.

Kendall said OEM manufacturers who put the advanced Wayve (ADAS) support system in new manufacturing vehicles should invest anything in additional hardware, as technology can work with existing sensors that usually consist of surround chambers and some radar.

Wayve is also “silicon agnostic”, which means that it can manage its software at any GPU, its OEM partners already have in their vehicles, according to Kendall. However, the current navy for the development of the starter uses the NVIDIA chip system.

“Entering ADAS is really critical because it allows you to build a sustainable business, build a scale distribution and get data exposure so you can train the system to (level) 4,” Kendall said on the stage on Wednesday.

(Level 4 driving system means that one can only navigate an environment – under certain conditions – without having to intervene.)

Wayve plans to first commercialize its system at the Adas level. So, the launch designs AI’s driver to run without Lidar – the detection of light and radar that measures the distance with laser light to generate a highly accurate 3D map of the world, which most companies developing level 4 technology considers a major sensor.

Weve’s approach to autonomy is similar to that of Tesla, which is He also works on a deep -to -ending model to power her system and constantly improve his self -government software. While Tesla tries to do, Wayve hopes to use an ADA widespread performance to collect data that will help its system reach complete autonomy. (Tesla’s full self -government software can perform some automated driving tasks, but not fully autonomous. Although the company aims to launch a Robotaxi service this summer.)

One of the main differences between Wayve and Tesla approaches from a technological point of view is that Tesla relies only on cameras, while Wayve is happy to include Lidar to reach a short -term complete autonomy.

“The longer -term plan is certainly the opportunity when you build the reliability and the ability to validate a level of scale to shrink this (sensory package) further,” Kendall said. “It depends on the experience of the product you want. Do you want the car to drive fog faster? Then you may want other sensors (like Lidar). But if you are ready for AI to understand cameras limit and be defensive and conservative as a result? Our AI can learn this.”

Kendall also annoys the GAIA-2, the latest Wayve generative model, adapted for autonomous driving that trains its driver in huge quantities of real and synthetic data in a wide range of tasks. The model processes video, text and other actions together, which Kendall says, allows Wayve’s AI driver to be more adaptable and like a person in his driving behavior.

“What is really exciting to me is the human driving behavior you see it appears,” Kendall said. “Of course, there is no hand -coded behavior. We are not telling the car how to behave. There are no infrastructure or HD cards, but instead the occurrence of behavior is managed by data and allows for driving behavior that deals with many complex and varied scenarios, including scenarios that he may never have seen during training.”

Wayve shares a similar philosophy with the autonomous start to Waabi transport, which also pursues an end -to -end training system. Both companies emphasize scaling managed by data AI models that can summarize through different driving circles both rely on Generative AI simulators test and train their technology.