On January 20, Xiaoma Zhixing unveiled for the first time the appearance design, sensor and computing platform solution of the sixth generation of autonomous driving software and hardware system design for L4 vehicle specification mass production. The first model to be equipped with the system is the Toyota S-AM, a hybrid electric platform based on the seven-seat Xena, which will start road testing in China this year and start the daily operation of the Autonomous Mobility Service (Robotaxi) in the first half of 2023.
The Koma Chi Heng's autopilot system for mass production is installed on the Toyota S-AM model
According to reports, the automatic driving system is aimed at vehicle-level mass production from the aspects of modeling design, parts research and development and selection, software and hardware coupling, safety redundancy and system assembly production. This is the result of Xiaoma Zhixing's continuous optimization of the performance and cost, appearance and practicality of the overall automatic driving solution since the beginning of the road.
At present, one of the important strategic goals of Pony Zhixing is to promote the application of autonomous driving technology in the field of passenger cars, for which Xiaoma Zhixing has accumulated more than 10 million kilometers of autonomous driving public road test mileage worldwide. Modular application landing. Peng Jun, co-founder and CEO of Xiaoma Zhixing, said that he will continue to increase technology investment and expand the scale of Robotaxi's operations. Toyota's S-AM model joining the PonyPilot+ team is a major result of the in-depth strategic cooperation between Xiaoma Zhixing and Toyota, and the new generation of automatic driving software and hardware systems designed for mass production of L4 vehicle rules will further improve technical reliability and promote the large-scale application of automatic driving technology.
He revealed that Xiaoma Zhixing and Toyota began to explore the feasibility of autonomous driving of the Chinese version of S-AM in 2019, and the two sides fully considered the actual scenarios and technical requirements of the application of automatic driving technology, so that the customized S-AM has a dual-redundant system and better control performance for L4 level automatic driving functions. In addition, as an MPV model, the S-AM has a flexible and practical seating space to meet a variety of use scenarios such as family travel.
Specific to the new generation of self-driving software and hardware systems disclosed this time, compared with the previous generation, it still follows the technical route of multi-sensor deep integration (Sensor Fusion), but the number of sensors and the selection of models are fully upgraded towards mass production, and the first large-scale use of solid-state lidar, the total number of sensors reaches 23, of which the resolution of self-developed signal light recognition cameras is increased by 1.5 times.
The sensor scheme consists of 4 solid-state lidars on the roof; 3 blindfolding lidars distributed on the left and right sides and backwards; 4 millimeter wave angle radars at the four corners of the roof; 1 forward long-range millimeter wave radar; and 11 cameras, 7 on the roof and 4 around the body.
At the same time, this set of mass production system solutions adopts the Xiaoma Zhixing self-developed vehicle regulation level L4 automatic driving calculation unit, and introduces the vehicle specification level NVIDIA DRIVE Orin system level (SoC) chip, which has higher performance, higher bandwidth and lower latency, and realizes vehicle regulation-level mass production through cooperation with the industrial chain.
Compared with the previous generation of computing units, the new generation is expected to increase the hash rate by at least 30%, reduce the weight by at least 30%, and reduce the cost by at least 30%. Combined with the experience of the previous generation of computing platforms, the water-cooled heat dissipation scheme has also been optimized and upgraded.
Based on the strategic cooperation and in-depth customization with NVIDIA, Xiaoma Zhidi has launched a variety of computing unit solutions with different chip configurations, based on one or more NVIDIA Orin system-level chips and vehicle-grade NVIDIA Ampere architecture GPUs, which can meet the technical requirements of the research and development of autonomous passenger cars and autonomous commercial vehicles at the same time, and accelerate the large-scale mass production of autonomous driving technology by Xiaoma Zhixing.
The new generation of systems has multiple layers of redundancy, including complete software, computing platform and vehicle platform redundancy system, to maximize the balance of scene coverage and safety processing mechanism. In an emergency, safe pullover or parking within the safe lane line can be achieved to minimize safety risks.
The new generation of sensor kits features a front-to-rear two-stage split design
This time, Xiaoma Zhixing first demonstrated the appearance of an autonomous vehicle equipped with a new system, which aims to reflect the modeling design concept that is closer to mass production, and make the perception module more integrated and beautiful. The new generation of sensor kits features a unique front-to-rear two-stage split design with one solid-state lidar in the front and three solid-state lidars in the rear, presenting a lighter visual experience and in line with aerodynamic and vehicle streamlined trends.
Xiaoma Zhixing also designed an innovative lighting language based on different scene uses, which not only adds visual reference to the operating state of various systems, but also helps Robotaxi users quickly identify pre-ordered autonomous vehicles. This will provide important design trend guidance for robotaxi in the future.
Text/Guangzhou Daily, New Flower City Reporter: Deng Li
Guangzhou Daily New Flower City Editor: Li Guangman