Perception Platform facilitates production of autonomous vehicles

March 29, 2021 // By Christoph Hammerschmidt
Perception Platform facilitates production of autonomous vehicles by major OEMs
Lidar and perception software expert Innoviz is shipping its Automotive Perception Platform to major automotive customers worldwide for evaluation. The InnovizApp provides these OEMs with access to critical technology for building and testing autonomous driving systems.

The InnovizAPP perception platform includes automotive-grade hardware and software and enables autonomous vehicles (AVs) to identify and classify objects. Innoviz is already supplying the product to selected car manufacturers, helping them bring their AV projects to market faster.

The platform is based on Innoviz's Perception software. It processes data from the manufacturer's lidar sensors in conjunction with proprietary AI algorithms to analyse the point cloud and estimate the speed of an object with high precision. The software can accurately detect and classify objects in any 3D driving scene up to 250 metres away. This allows InnovizAPP to distinguish and classify cars, trucks, motorbikes, pedestrians and other objects. It also runs perception algorithms in real time that detect and classify pixels as collision-relevant or non-collision-relevant.

The InnovizAPP hardware is a reference design based on low-cost automotive-grade components. It connects to existing vehicle systems and enables real-time perception in a simple plug-and-play manner, allowing car manufacturers to test and learn how to build their own autonomous driving systems.

Both Innoviz InnovizOne's perception software and solid-state lidar sensor are designed and developed to meet the highest automotive industry standards and are already being used by major players such as BMW. 


Related articles:

Samsung, SoftBank back lidar startup Innoviz

Magna-Innoviz will supply Lidar sensors to BMW


Vous êtes certain ?

Si vous désactivez les cookies, vous ne pouvez plus naviguer sur le site.

Vous allez être rediriger vers Google.