Feature April
Photo Courtesy: Amap & AutoX

China AV - Apr. Newsletter

April had been a challenging month to the global community and the automotive industry. OEMs dedicated their resources and expertise to manufacture ventilators and save lives in their countries, while 3 billion people entered lockdown. While we all learned how to remote work effectively with our teams, the ecosystem at China is already back on track pushing forward. Politics aside this is a great opportunity for global companies that are interested to tap this market to enter and collaborate with local partners to leverage the rapid internal growth.

During April both Baidu and AutoX launched 2 new RoboTaxi services, while Baidu leverage their popular mapping app to book rides directly with final users. This move emphasizes the dependence of autonomous vehicle providers on social networks to access the final customer.

Apr.1st: DeepRoute.ai (元戎启行), a L4 autonomous driving provider, published its latest findings at CVPR 2020. In order to ensure autonomous driving safety, the most important task of 3D object detection is to balance between accuracy and time-consuming, so that the system can quickly identify and respond in real time. Taking this equation into account, the company proposed a new point cloud-based unified network model for 3D object detection using a mixed voxel network named HVNet: Hybrid Voxel Network, through mixing scale voxel feature encoders (VFEs). Furthermore, in order to maximize HVNet capabilities the team developed an inference engine, which optimizes computing resources for HVNet’s custom operators and network structure.

deeproute
deeproute2

Apr.3rd: QCraft.ai (轻舟智航) had raised a seed funding round from IDG Capital, Vision+ Capital, and Tide Capital. QCraft is building a virtual driving system (VDS) focused on challenging situations in complex urban environment. It is heavily using simulation system and a self-learning framework for decision-making and planning.

Qcrafti

Apr.7th: Qianxun SI (千寻位置) has obtained IEC61508 and ISO26262 functional safety certification, becoming the only high-precision positioning service provider in the world that has passed both certification.

Qianxun SI

Apr.7th: A new Leaderboard Report from Navigant Research examines the strategy and execution of 18 companies developing automated driving systems. Baidu was ranked 4th in the global ranking. Important to mention that several of the top 10 list are collaborating between each other. For example, both Daimler and Ford are part of Apollo-Baidu program.

Top 10 Vendors

  1. Waymo
  2. Ford Autonomous Vehicles
  3. Cruise
  4. Baidu
  5. Intel-Mobileye
  6. Aptiv-Hyundai
  7. Volkswagen Group
  8. Yandex
  9. Zoox
  10. Daimler-Bosch

Apr.9th: Baidu Apollo officially released “ACE Transportation Engine“, which disclosed the Apollo intelligent transportation solution for the first time, and reveled Baidu’s business plan for its developments at artificial intelligence, autonomous driving, and vehicle-road collaboration.

Apr.9th: Huawei showed the HMI operation screen of the HiCar mobility solution for the first time. Compared with traditional solutions such as Android Auto and CarPlay, HiCar can achieve deeper integration with vehicles. Huawei HiCar is collaborating with over 30 car manufacturers, including Audi, FAW, GAC, BAIC, Chery and JAC.

Huawei

Apr.14th: China FAW Group established wholly owned FAW Nanjing Technology Development company. The new group business scope includes artificial intelligence software development, algorithm software development for AV and intelligent vehicle equipment manufacturing.

Apr.15th: NavInfo (四维图新) released 2020 Q1 estimated results, which were heavily affected by COVID-19 pandemic. Net loss is RMB 59M to RMB 84M, a sharp decline compared with 2019 Q1 profit RMB 42M. Domestic car sales decline during the first quarter, resulting in decrease at company’s mapping sales revenue and a sharp decrease in the automobile electronic chip shipments. March shows significant improvement in the domestic car market performance, as for the gradual releaf in the epidemic situation. Overall production and sales volumes are still extremely low, production is down 44.5% compared to 2019Q1 (total 1.422 million vehicles) and sales are down 43.3% (1.43 million vehicles).

Apr.16th : Xpeng Motors announced that it will continue to cooperate with NVIDIA during mass production. The Xpeng P7 will be the first mass-production model equipped with the NVIDIA DRIVETM AGX Xavier platform, which can achieve 30 trillion operations per second and support L4 autonomous driving computing power requirements. NVIDIA DRIVETM AGX Xavier consumes 30 watts – 15 times! more energy efficient than the previous generation architecture.

Apr.19th: Baidu fully opened its robotaxi service at Changsha, enabling users to hail an autonomous taxi free of charge through the Baidu Map app and Baidu’s “Dutaxi” app. The service is available at an area of 130 square kilometers and its driving route covers various life scenarios such as residential areas, business and leisure areas, and industrial parks in Changsha.

Apr.21st: An autonomous L3 heavy truck jointly developed by Inceptio.ai and Dongfeng has successfully completed the acceptance of A-type truck. The evaluation includes three rounds: static evaluation- 16 items in 4 categories, dynamic evaluation- 29 items in 6 categories, and autonomous driving function evaluation on a highway. This is the first time at China a prototype vehicle had completed the AV L3 production development process.

Apr.22nd: Ali Damo Academy released the world’s first “Hybrid simulation and test platform” for autonomous driving, based on simulation technology that combine virtual and reality. The platform introduces real road test scenes and cloud trainer. It only takes 30 seconds to simulate an extreme scene, what allows the daily virtual test mileage of the system to exceed 8 million kilometers! These capabilities can rapidly improve training efficiency of a virtual driving systems before implementing those at autonomous cars.

Apr.25th: Chinese urban rail transit signaling system provider, FITSCO (Shanghai Fuxin Intelligent Transportation Solutions Co) and Russia’s Cognitive Pilot, joined forces to develop an AI-based computer vision system for urban trams featuring autonomous control and movement. The system main target is to increase safety and efficiency of public transportation. “The new technology will reduce the number of accidents in the city and will minimize the dependence on the human factor”, says Jack Wu, CEO of FITSCO.

Apr.27th: AutoX has partnered with Amap mobility platform by Alibaba Group, in order to offer the people of Shanghai an option to hail a ride in an AutoX RoboTaxi. The plan is to deploy 100 RoboTaxis on public roads at Shanghai, which will all communicate between each other using road infrastructure via the city’s 5G-based V2X technology to gather traffic data.

Apr.30th: BYD officially launched DiPilot, a brand-new advanced driver assistance system (ADAS). It is reported that the new ADAS will be applied to the BYD Han first. DiPilot team is focused on the development of big-data algorithms of DiTrainer, which can learn the driver’s driving habits, judge the driver’s style and driving skills, and optimize the ADAS for the user.

Sources: Toyota; Leiphone; QCraft.ai; Gasgoo; ChinaNews; Bloomberg; Guidehouse Insights; cnTechPost; Pandaily; Carscoops; New Atlas

Leave a Reply