Momenta Rolls out Mpilot Parking Solutions
Chinese autonomous driving company Momenta.ai released its self-parking solutions Mpilot Parking on July 11. The in-dash product is designed to replace average 30-minutes that the driver spends on parking while improving road safety.
Chinese autonomous driving company Momenta.ai (初速度) released its self-parking solutions Mpilot Parking on July 11. The in-dash product is designed to replace average 30-minutes that the driver spends on parking while improving road safety, according to Momenta. After applying the solutions, the driver only needs to click on a button and choose a parking space and the system works automatically, according to a video viewed by iyiou.com.
HD Map is essential to the autonomous driving system. Momenta’s HD Map solutions is able to provide localization for autonomous vehicles and map updates through crowdsourcing. Through crowdsourcing, the solutions can create a closed feedback loop of big data, AI and HD map updating.
On the basis of localization, Momenta will discover changes in the map elements and provide frequent updates to the cloud. In terms of HD map, the firm has one strong partner, Ambarella, Inc. (NASDAQ: AMBA), a Santa Clara, California-based developer of high-resolution video processing and computer vision semiconductors.
Based on the HD map for the parking lot that is previously uploaded, the company adopted a set of sensors to sense surroundings, including four Fisheye surround view cameras, One front view camera, 12 ultrasonic sensors, one consumer-grade IMU (Inertial Measurement Unit), and one consumer-grade GPS. Leveraging SLAM (self-localization and mapping) and vision sensing (perception of lanes, traffic signs, and other objects), Momenta’s deep-learning software algorithms could be engaged efficiently on the automotive-grade chip.
The firm cooperates with semiconductor player NXP for developing automotive-grade DMS (Driver Monitoring Solutions). NXP’s automotive-grade hardware accelerators integrated in the NXP S32V2 are perfect for deep neural network processing as they can save computing resources and reduce CPU usage.
The product only costs thousands of yuan, the company claimed. For the record, one Lidar, a commonly used (for example, Alphabet’s Waymo) sensor for Level 4 autonomous driving is priced at thousands of U.S. dollars.