Recently, Autowise.ai launched the fourth generation of self-developed AW Roboard-X. This product can realize multi-scene upper X transformation, and adopts cab-less design with high-tech appearance. The birth of AW Roboard-X means that Autowise.ai autonomous vehicles can fully support efficient reuse in different scenarios, demonstrating the ability of technology migration, optimization and iteration of Autowise.ai.
Roboard is the abbreviation of Robot-board, which is different from the traditional skateboard. Roboard is a robot chassis with fully autonomous system. X stands for diversified upper X. It means the autonomous vehicles can be used for passenger vehicle, cargo transportation and sweeping, to achieve multi-use and modular development of one vehicle.
Since its establishment in 2017, Autowise.ai has continued to deepen its efforts in autonomous driving industry. After the implementation of passenger car and commercial vehicle technologies, Autowise.ai has taken the lead in commercializing autonomous driving technology in the field of urban sanitation and became a leading company in the commercialization of autonomous driving.
The development of autonomous products from Autowise.ai have gone through four stages: passenger cars, commercial vehicles, autonomous sweepers V3 and Roboard-X. Those reliable products have complete delivery experience in each stage, with the closed-loop productization capability from equipment selection, adaptation, autonomous driving system development to assembly, debugging and verification.
Autowise.ai autonomous vehicle evolution path
The perception equipment of the first-generation passenger car was relatively single, with lower computility. The second-generation of commercial vehicle perception equipment coverage was significantly improved. The computility and electrical system were gradually upgraded to Automotive Grade, and the equipment tended to be stable. The whole vehicle and system of the third-generation autonomous sweeper V3 were independently developed by Autowise.ai. V3 has a strong sense of science and technology, a perception of 360°coverage. The drive-by-wire was Automotive Grade and the system was reliable and highly responsive. The computing module switched into SOC, with effective heat dissipation and stability.
After the exploration, productization and commercial iteration of the previous three generations, the fourth-generation of self-developed AW Roboard-X not only realizes the complete separation of Roboard and upper X, but also upgrades the software and hardware technologies, such as perception equipment, wire control system and computing module. In terms of perception, the reliability and accuracy of obstacle detection are significantly improved through the fusion of traditional algorithm and deep learning algorithm, especially the fusion of LiDAR, camera and millimeter wave radars in BEV perspective. Meanwhile, the data augmentation method based on multi-view consistency (published in CVPR2022) and semi-supervised training framework (published in ECCV2022) can significantly reduce the dependence on the amount of labeled data while ensuring the same accuracy, thus helping the continuous improvement of perception performance with lower R&D cost. For the vehicle, Automotive Grade EHB and SBW are used for high response accuracy. The whole system adopts Dual Redundant, which makes the system stable and reliable. The packaging interface is open and modular, which can realize quick loading. The computing module switch to Automotive Grade SoC, dual-Orin, with AI performance over 600TOPs, redundant and stable.
The perception scheme and coverage of AW Roboard-X, 360° full coverage of surroundings
At present, AW Roboard-X of Autowise.ai has realized the X of passenger vehicle (AW Roboard-Space), logistics vehicle (AW Roboard-Van) and sweeper (AW Roboard-Sweeper), with a cab-less design. The three autonomous models of passenger vehicle, logistics vehicle and sweeper have been implemented for mass production, which is practical and technological. Roboard-X can meet the requirements of different scenarios in terms of hardware and software architecture design.
Yanye Tang, Vice President of Product and Operation, said: “The Roboard and X is developed separately. Roboard adopts a platform model, the motor, battery and thermal management system are modularized. The upper X can be subsequently iterated independently, thus the development cycle and efficiency will be greatly improved. The application of Roboard-X can realize the cost reduction of both user side and R&D.”
Mario Pepe, Head of Sales in EMEA region, said: “Roboard-X fits perfectly with all my imagination for the future autonomous smart space. Its leadership is not only reflected in the diversity and convenience of model development, but the platform and modularity of the manufacturing and supply chain will also enable autonomous driving vehicle to step into a new era of mass production. Roboard-X truly breaks the product boundary of autonomous driving.”
With the release of the self-developed AW Roboard-X, Autowise.ai will continue to further broaden the application scenarios of autonomous vehicles with its strong technology strength and the rich experience of commercialization. Autowise.ai will continue to promote the commercialization of autonomous driving.