Black Sesame Technologies’ computing solution deployed in smart roadside infrastructures

Beijing (Gasgoo)- Recently, smart computing solution developer Black Sesame Technologies participated in the launch of an autonomous driving freight logistics corridor across Beijing, Tianjin, and Hebei regions.

Roadside unit demo; photo credit: Black Sesame Technologies

The automated freight vehicles now operate seamlessly along the Beijing-Tianjin-Tanggu expressway. At the Majuqiao toll station merging area towards Tianjin, Black Sesame Technologies delivered a comprehensive holographic ramp solution. This marks the first instance of autonomous vehicles making lane-changing decisions based on roadside sensing data in open road conditions, signifying Black Sesame Technologies' official entry into the integrated vehicle-road-cloud application development for intelligent connected vehicles, the company said.

Since 2019, Black Sesame Technologies has recognized the value of localized high-computing power chips in roadside edge computing. Aside from its primary focus on high-performance automotive-grade computing chips, the company has ventured into a secondary field, providing highly reliable high-computing power chips for roadside sensing, computing, and applications to intelligent transportation system integrators and equipment suppliers.

After five years of collaboration with ecosystem partners on both the vehicle and road sides, Black Sesame Technologies has developed three major product lines centered around its A1000 chip, including intelligent driving domain controllers and vehicle-side perception algorithms, edge computing units and roadside perception algorithms, and vehicle-road integrated perception algorithms, forming a full-stack vehicle-road capability.

The A1000 chip, compliant with all automotive-grade certifications and suitable for mass production, supports integrated driving and parking domain controllers. With a single-chip computing power of 58 TOPS (INT8), it features extensive sensor interfaces, supporting L2+/L3 autonomous driving solutions. It integrates CPU, ISP, DSP, and NPU to meet various computing needs.

The edge computing all-in-one machine equipped with the A1000 chip supports up to 12 cameras and 8 millimeter-wave radars for parallel analysis, outputting over 200 parameters across 63 indicators in four categories: target detection, trajectory tracking, traffic parameter detection, and traffic event detection, with a computing power utilization rate of less than 50%, ensuring stable operation and allowing for future expansion into intelligent traffic control and large model applications.

With these capabilities, Black Sesame Technologies and its system integrators can deliver high, medium, and low-end sensing solutions for holographic road networks, intersections, and expressways, achieving large-scale deployment in multiple cities.