QCraft’s new products reduce AD solution cost to RMB10,000

Beijing (Gasgoo)- On May 18th, China’s autonomous driving developer QCraft introduced three core products, namely Longzhou Space, production-ready autonomous driving solution DBQ V4, and autonomous driving R&D toolchain QCraft Matrix, at the first QCRAFT DAY. 

Photo credit: QCraft

The DBQ V4 is QCraft’s fourth-generation mass-produced autonomous driving solution, integrating its full-stack homegrown software and hardware technologies. The solution offers stellar cost performance, realizing 99% of L4 functions with only 10% of the normal cost. The mass production cost of the solution is reduced to RMB10,000 ($1,483).

The solution can be installed with up to 5 LiDAR units, 4 blind-spot radars, 6 millimeter-wave radars, and 12 perceptual cameras, providing a stable 360° redundant perception under various light conditions and vehicle moving status. The solution is also equipped with a customized camera that recognizes traffic lights in shape and colors.  

The DBQ V4 is completely localized, and customizable to cater to automakers’ different needs in configuration and vehicle models. The basic configuration scheme of the solution shares the same technology stack, sharing and collecting data seamlessly. 

The Longzhou SPACE is a mobility space solution dedicated to complex open road operations. The solution is adaptive to various vehicle genres, and trip distances, with changeable and flexible space to meet the demands of different scenarios. 

In December 2021, QCraft has deployed the Longzhou SPACE solution onto its Sharing Bus co-developed with Dongfeng Motor. The buses are in operation in Wuhan, Dali, and other cities. 

Last but not least, QCraft created its own autonomous driving R&D toolchain based on simulation, the QCraft Matrix. The toolchain has started servicing clients, in order to feedback to its autonomous driving super factory.

The QCraft Matrix breaks through the entire flow scheme from data processing, annotation, training, large-scale simulation, and technical output. The simulation scenarios building on real-life road test data can reduce autonomous driving test costs to 1% of real road tests. The solution can generate millions of long tail scenarios, completing millions of daily algorithm training, test validation, and iteration.

The QCraft Matrix has been opened up to external clients. In an effort to accelerate large-scale commercialization of autonmosu driving technologies, QCraft has partnered with ByteDance’s Volcengine, T3 Mobility, and Horizon Robotics.