Green Hills Software and StradVision Demonstrate Road-Ready Deep Learning Perception at CES 2020
Safe and Secure INTEGRITY RTOS and StradVision’s SVNet Vision Processing Technology Combine for Deep Learning Driving in Late 2020 Vehicles
When the biggest stakeholders in the automotive and technology worlds converge in Las Vegas at CES 2020 from Jan. 7-10 to showcase the latest in software developments, Autonomous Vehicle technology will be front and center.
One noteworthy collaboration at CES 2020 is between StradVision, an innovator in deep learning-based autonomous vehicle vision processing, and Green Hills Software, the worldwide leader in embedded safety and security.
In a joint technology demonstration, StradVision’s SVNet software will leverage the separation and resource protection architecture of the Green Hills INTEGRITY® real-time operating system to safely and reliably execute its deep learning-based embedded perception algorithms.
The CES demonstration showcases the collaboration between the two software companies on systems for Autonomous Vehicles currently in development, including a front-camera project with a leading global OEM and Tier 1 supplier for vehicles on the road in late 2020.
“With autonomous vehicles comes a tremendous responsibility to offer software that is safe and reliable,” said Green Hills Software’s Vice President of Advanced Products, Tim Reed. “The combination of INTEGRITY with StradVision’s SVNet embedded perception algorithm will allow drivers of autonomous vehicles to feel confident in the performance and safety of their autonomous vehicle in real-world scenarios.”
StradVision CEO Junhwan Kim said partnerships like the one with Green Hills are critical, as collaboration is an essential component of getting autonomous vehicles to the mass market.
“As StradVision deploys its SVNet software to millions of autonomous vehicles around the world in the years to come, collaboration with other leaders in this growing space will be essential,” Kim said. “Partnerships with respected and proven brands in the AV sector like Green Hills will play major role in giving drivers the peace of mind they demand.”
To achieve the highest levels of safety and security, Green Hill Software’s INTEGRITY real-time operating system was designed so embedded developers could ensure their critical applications are protected from malfunctions and malicious code and have access to the necessary CPU and memory resources to successfully run in a reliable and deterministic manner. For over 20 years, INTEGRITY has controlled the most critical of applications across the spectrum of embedded systems – from aircraft flight computers, to surgical devices, to automated driving functions. As such, it has achieved the highest levels of automotive safety certification—ISO 26262 ASIL D.
The Renesas Electronics Corporation is an important partner in the CES demonstration because the automotive-grade R-Car V3H System-on-Chip provides hardware platform and acceleration units, utilized by INTEGRITY and StradVision’s SVNet software.
StradVision’s deep-learning expertise has helped the company’s algorithms lead the way in Object Detection (OD) and Free Space Detection (FSD), and its SVNet External software enables vehicles to execute ADAS and self-driving functions, including:
- Automatic Emergency Braking
- Adaptive Cruise Control
- Lane Departure Warning
- Blind Spot Detection
StradVision, which recently was honored with a 2020 AVT ACES Award, offers a wide range of abilities in the autonomous realm 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 ADAS systems and autonomous vehicles.
By 2021, StradVision will have millions of vehicles on the world’s roadways using SVNet, which is already compliant with strict standards such as Guobiao (GB) in China. StradVision is continuously expanding in China, and already deploying ADAS vehicles on Chinese roads. StradVision currently has multiple production projects in Europe and China, including work on a Level 4 autonomous bus project in Europe.