American start-up company NPS (Neural Propulsion Systems) announced on March 5, 2021 that the company will launch the sensor platform "NPS 500", which integrates and encapsulates various sensors essential for autonomous driving. It is said that it can be used for L4 Or higher autonomous driving.
NPS was founded in 2017 by entrepreneur Behrooz Rezvani (current CEO) and other partners and is headquartered in Pleasanton, California. Before establishing NPS, Rezvani was the founder of semiconductor provider Ikanos Communications, wireless communication technology development Quantenna Communications and Seyyer. He holds more than 40 patents in the United States and specializes in electrical engineering and quantum devices.
NPS 500 integrates solid-state MIMO-LiDAR, ultra-high resolution SWAM radar and camera. It is an all-in-one packaged product that covers all the sensors necessary for autonomous driving. According to the company, this is the world's first packaged product that integrates these three sensors at the same time.
Its innovative architecture can double the detection range and resolution of MIMO-LiDAR, the detection reliability of radar is 10 times higher than that of existing products on the market, and AI fusion technology can combine the data and resolution of ultra-high-resolution cameras. Other sensors are combined to make up for the shortcomings of each sensor.
Specifically, it can detect objects 500 meters away with a reflectivity of 10%, pre-detect pedestrians and other vehicles at intersections and predict behavior, and detect vehicles that may not conform to traffic rules, and can also detect road vehicles. Pedestrians blocked by trees on the side and vehicles in view.
Generally, self-driving cars will be equipped with 10 or more various sensors, including LiDAR, cameras, millimeter-wave radar, etc., but the configuration of each is different, because each company has different ideas about the construction of sensor systems.
In this case, the importance of sensor fusion (Sensor Fusion) technology is increasing. For example, how to effectively integrate the information collected by each sensor and enhance the recognition capabilities, which is very important for car manufacturers and first-tier suppliers.