Onsemi's Hyperlux image sensor will be integrated into the NVIDIA DRIVE platform, enabling accurate and smart decision-making and rapid responses in autonomous vehicles.

ON Semiconductor, a leader in intelligent power and sensing technologies, today announced that its Hyperlux™ image sensor family is now supported on the NVIDIA DRIVE platform, enabling enhanced vision and improved safety for autonomous vehicles.

This powerful combination of technologies enables autonomous driving systems to leverage the full potential of their sensors with superior image quality that captures more detail under all sensor conditions.

One of the most crucial functions in autonomous driving is the vision system. The Hyperlux image sensor delivers exceptionally detailed images without the need to adjust exposure settings, eliminating blind spots. It also provides the highest quality data, enabling the NVIDIA DRIVE platform to quickly process and make informed decisions. Combined with the efficient architecture of NVIDIA platforms, these sensors deliver industry-leading low power consumption, making them the lowest power option available for autonomous vehicle (AV) operation.

NVIDIA Drive is a scalable, open AI platform comprised of hardware, software, and firmware that work together to enable the production of automated and autonomous vehicles.

“Self-driving cars require exceptional image quality, and the vehicle must outperform the driver,” said Chris Adams, vice president of Automotive Sensing at ON Semiconductor. “By offering Hyperlux as part of the NVIDIA DRIVE platform, automakers can gain a system that can detect problems before the driver does and make decisions to avoid them. This is a critical step toward achieving fully autonomous driving.”

The Hyperlux image sensor is the only sensor currently available that meets these expectations while delivering 150dB HDR. For example, under direct sunlight, which can blind a human, the Hyperlux image sensor enables autonomous vehicles to distinguish between yellow and red traffic lights from a distance. Even in extremely dark environments, its super-exposed pixel technology exceeds requirements, capturing people, animals, and objects that are otherwise difficult to see while driving.

The Hyperlux sensor provides the NVIDIA DRIVE platform with the information it needs to make more accurate decisions. This prevents other sensors from having to change exposure modes for sharper images or from missing frames that could contain important details. Therefore, these delay-free systems can see and make decisions about everything in their field of vision, including objects invisible to the human eye.

“Building public trust in autonomous driving means ensuring that supercomputers on wheels can safely get users where they want to go,” said Glenn Schuster, senior director of sensor ecosystems at NVIDIA. “With Hyperlux image sensors now part of our ecosystem, we can continue to offer OEMs and tier-one customers the best choice to meet their needs, ensuring the highest levels of system performance, functional safety, and cybersecurity.”

To ensure safety, image sensors must operate reliably under all conditions. Extreme temperatures degrade image quality, reducing the efficiency of the entire system for recognition tasks such as object detection and classification.

The Hyperlux family performs to specification and delivers best-in-class image quality for both viewing and sensing applications across the full vehicle temperature range from -40°C to +125°C, preventing performance degradation.

The Hyperlux family not only delivers sharp images even at high temperatures, but also enhances system safety by capturing fast-moving objects, car headlights, and flashing lights.