Neural Propulsion Systems (NPS), founded in 2017 by Silicon Valley luminaries, specialise in hyper-resolution digital imaging radar technology. Their AtomicSense radar technology been recognised for its exceptional performance—it won a CES 2023 Innovation Award.
AtomicSense offers faster scanning time, 10× better resolution, and a 2.5× improvement in range compared to conventional radar offerings. This performance is made possible by NPS’ innovative algorithms, based on frequency modulated continuous wave (FMCW) and array processing.
The AtomicSense platform is described by NPS as the world’s first all-in-one, deeply-integrated, multimodal sensor system for ADAS and AVs at high volume and affordable cost. The AtomicSense components fuse data from the NPS solid-state MIMO lidar; super resolution SWAM radar, and cameras to deliver consistent and accurate, high-resolution, 360° data. The system is powered by a new class of estimation and detection algorithms based on the Atomic Norm, and has shown terrific performance advantages in excess of 50× over current solutions. The densely integrated deep sensor-fusion system gives vehicles the ability to see around corners and over 1,000 metres away.
The association of a lidar and an imaging radar can improve the reliability of this global perception system, especially in adverse weather conditions. Camera perception performance in fog; spray; rain, and snow is something of a pain point or bottleneck on the way toward widespread autonomous driving, and a great deal of work is being done in this area.