Momenta, which is building the “brain” for autonomous driving, has partnered with Toyota to provide automated HD mapping and updates through vision-based technologies.

To better serve Chinese customers, the companies will seek to use this joint development to promote the commercialization of Toyota’s Automated Mapping Platform (AMP) in the Chinese market.

Momenta’s vision-based, camera-based HD mapping uses low-cost consumer-grade sensor sets, which is made up of camera, GPS, and IMU to automatically generate HD maps with 10-centimeter level relative accuracy by using cutting-edge technologies such as deep-learning based perception and simultaneous localization and mapping (SLAM).

Momenta’s HD map includes geometry features such as traffic lights and signs, poles and road markings. The map also generates road level and lane level topology and semantic features.

Momenta says this technology has generalization capability of a number of different devices, which is why it is available for large-scale cost-effective commercial applications.

Through automated mapping pipeline, the HD map processes big data, allowing it to be updated with high frequency. This results in a “live,” reliable map for different autonomous driving modules such as localization, planning and control.

“Momenta and Toyota’s cooperation will make the best of two companies’ strengths,” Momenta says.

“Momenta’s strategy is to pursue both mass-production passenger vehicles and full autonomy for mobility service vehicles. We are confident that the cooperation with Toyota will play an important role in this strategy.”