StradVision, a South Korean company creating vision processing technology for self-driving vehicles, has announced a partnership with a leading global Tier 1 automotive supplier to develop custom camera technology for SAE Level 4 autonomous buses.
StradVision says Level 4 autonomous buses will be among the key innovations leading the way to the widespread use of self-driving vehicles, and the company’s versatile SVNet software can play a major role in their success. Once production is completed in a few years, the technology will be implemented by multiple OEMs in tens of thousands of Level 4 autonomous buses that will operate in French and German towns within a predetermined ‘geofenced’ border, without any human supervision. The project, for which production began in September, will focus on three key areas: Object Detection (OD), Traffic Sign Recognition (TSR), and Traffic Light Recognition (TLR).
StradVision’s wide range of abilities in the autonomous realm are possible due to the leaner structure of the SVNet network, as well as its flexibility. There are multiple types and iterations of SVNet, increasing its versatility for advanced driver-assistance systems (ADAS) and autonomous vehicles. The company’s SVNet software is designed to function properly even in poor lighting or weather conditions, and features three key components:
- SVNet External – object detection and recognition for vehicle perception;
- SVNet Internal – Driver Monitoring System (DMS);
- SVNet Tools – automated labelling and annotation.
For the autonomous bus project, SVNet External will be the key element, allowing the vehicles to safely and accurately navigate roadways when in self-driving mode.
SVNet can be used by low-end hardware to enable Level 2 ADAS functions, and also is capable of partnering with powerful hardware such as Nvidia’s Xavier system-on-a-chip processor, which is being used in the Level 4 autonomous bus project. Many global Tier 1s and OEMs are also finding the software to be a strong fit for their projects due to its compliance with key reliability certifications such as Guobiao (GB) in China, Automotive SPICE Level 2 and ASIL B. StradVision says that by 2021 there will be millions of vehicles on the world’s roadways using its SVNet software, with it already deployed in ADAS-equipped vehicles running on Chinese roads.
“In addition to our ability to provide vision processing systems for Level 2 ADAS solutions, we are also excited about our contributions to higher-level autonomous vehicles,” said StradVision’s CEO, Junhwan Kim. “This Level 4 bus project in the European market is one we’re particularly proud of, as it will show the world our ability to deliver truly cutting-edge technology with real-world applications, and the flexible nature of our product at multiple levels of autonomy.”