on 03 November 2023
The Open Atom Open Source Foundation launched the first “Open Atomic Open Source Competition” to unite open source organizations, enterprises and institutions, universities, research institutes, industry organizations, investment and financing institutions and other resources to give full play to the upstream and downstream of the industrial chain ecology. Collaborative capabilities, jointly organized based on the principles of open source sharing, co-construction and co-governance.
The competition builds a platform for cutting-edge technology competition, outstanding talent selection, innovation achievement display, business transformation guidance and docking exchange and cooperation platform for the global open source field, widely disseminating open source culture, popularizing open source knowledge, promoting open source projects, improving open source skills, and promoting the prosperity of the open source ecosystem. and provide motivation and support for sustainable development.
CARSMOS International Autonomous Driving Algorithm Challenge (hereinafter referred to as: CARSMOS Challenge) is the autonomous driving algorithm competition of the first Open Atomic Open Source Competition. It is hosted by the Ministry of Industry and Information Technology, the People’s Government of Jiangsu Province, and the People’s Government of Hunan Province, and is sponsored by the Open Atom Open Source Fund. Hosted by the OpenAtom Foundation, CCTV, Jiangsu Provincial Department of Industry and Information Technology, Wuxi Municipal People’s Government, Jiangsu Software Industry Talent Development Foundation, Suzhou Industrial Park, Wuxi High-tech Zone, etc., CARSMOS Yuanyao Open Source under the Open Atom Open Source Foundation The core units of the smart travel project team are Synkrotron Technologies and Futurewei. Technical supporters of the competition include the internationally renowned autonomous driving open source simulation community CARLA, Shanghai AI Laboratory, Tsinghua Mars Lab, OpenEV, etc.
The CARSMOS Challenge is open to the world. Excellent teams and players from domestic and foreign universities, research institutes, open source organizations, and industry organizations can register to participate. The challenge is an online competition, and the CARSMOS open source code repository (https://github.com/carsmos) is provided as a baseline reference design and implementation (not a mandatory condition).
The competition is based on a simulation environment, providing participating teams with multiple predefined training scenarios for learning and debugging, and then using more complex test scenarios to evaluate and compete each participating team’s autonomous driving algorithm in various complex traffic environments on highways and urban areas. perception, planning and decision-making abilities. The online competition will be ranked according to the online rankings, and the winning team will participate in the final online review meeting to determine the final winner.
Open registration and online exercise development/training: Users sign up, become familiar with the development and submission environment, and through online learning and training, can use the autonomous driving simulation platform Oasis and Oasis provided by the CARSMOS open source intelligent driving project (code repository: AtomGit) The middleware Dora is used to quickly develop a complete set of autonomous driving algorithms and can submit the algorithm to the system platform. At this stage, public training scenarios are used and no ranking is performed.
Formal online ranking competition: multiple predefined scenarios are provided in a simulation environment to test the perception, planning and decision-making capabilities of each participating team’s autonomous driving algorithm in various complex traffic environments. The scoring rules are transparent, and the algorithm competition platform automatically scores. At this stage, non-public test scenarios are used, and the list is made public.
During the defense and award stage, 1st, 2nd, 3rd prizes and excellence awards will be selected, and an award ceremony will be held; the winning team will receive generous rewards from the foundation and related internship and job opportunities.
1) Algorithm Open Source – Among the global autonomous driving competitions, the CARSMOS Challenge is the only open source challenge, which requires participating teams to open source their algorithms within 3 months, which will help future generations continue to improve and make the competition stronger and stronger.
2) The competition is of high standard – hosted by the Ministry of Industry and Information Technology, the People’s Government of Jiangsu Province and the People’s Government of Hunan Province.
3) Low entry barrier – CARSMOS has provided a very modular benchmark design, which is very convenient for participating teams to replace any operator module or subsystem to test their innovative algorithms.
4) Large talent reserve – The CARSMOS Challenge aims to attract and train college students/graduate students to enter the field of autonomous driving and cultivate a large number of reserve talents. The winners are outstanding engineers and future leaders that leading companies are competing to recruit.
The Yuan Ao CARSMOS open source smart travel project group (please refer to the English official website https://www.carsmos.ai/; the Chinese official website https://www.carsmos.cn/) is a strategic open source project under the OpenAtom Foundation. The group is led by Sangxin Technology, Futurewei, Arctic Xiongxin, and Open Motors. Currently, Changan Automobile, China Automobile Intelligent Control, Tsinghua Cross Information Research Institute and other companies and institutions are participating, and it is still developing and growing.
The project has a technical steering committee composed of world-renowned experts, including Li Keqiang, academician of the Chinese Academy of Engineering and professor of Tsinghua University, Guan Xiaohong, academician of the Chinese Academy of Sciences and Xi’an Jiaotong University, Yang Qiang, academician of the Canadian Academy of Sciences and Engineering and professor of the Hong Kong University of Science and Technology, and the National Academy of Sciences and Engineering. Academician and Professor Moshe Vardi of Rice University and others.
The Yuan Ao CARSMOS project group has six open source software and hardware sub-projects, including the autonomous driving open source application and community JUNO open source project, the autonomous driving and robot operating system middleware DORA open source project, and the vehicle-road collaborative edge computing service and management platform ESSA. Open source projects, autonomous driving simulation test platform Oasis open source project, intelligent electric vehicle design EDITEV open source project, and ChipletD2D interconnection interface standard Hermes open source project for NPU computing chips in the automotive field.
The Yuan Ao CARSMOS open source smart travel project group is committed to building a unified and open smart car hardware, computing, software and ecological platform, enabling smart driving open source basic software and hardware framework solutions, creating industry standards, enriching the industry ecology, cultivating industry talents, and promoting Industry-university-research cooperation progresses.
The mission of the Yuan Ao project group includes:
1) The world’s first automotive open source project community builds an open source ecosystem, gathers leading technologies, and moves from the era of logic-driven development 1.0 to the era of data-driven development 2.0;
2) China’s contribution to the development and achievement of the smart car industry software ecosystem;
3) Cultivate algorithm, software, and hardware talents that keep pace with the development of the automotive industry;
4) Accelerate the industrialization from innovative technologies to commercial applications.
2) Download the autonomous driving simulation system Oasis.
3) Be familiar with the autonomous driving simulation system Oasis and develop your own algorithms by referring to the documentation.
4) Refer to the document submission algorithm and wait for the running results after successful submission.
5) Competition training video collection.
Registration for the competition has been opened, and it is currently in the practice stage. You can participate in the competition now, get familiar with it first, and have a better chance of achieving better results and winning more generous rewards, with bonuses up to 500,000!
Please scan the QR code to join the CARSMOS Challenge Service WeChat group. If the group has exceeded the number of people, you can add the WeChat ID of Synkrotron Enterprise Customer Service (ID: synkrotron1) and remark “Contest” to join.
If you encounter any problems during the competition, you are welcome to fill in the “Carsmos Open Source Autonomous Driving Algorithm International Challenge·Problem Collection Form” and provide feedback.
Synkrotron is a national high-tech enterprise focused on providing autonomous driving simulation and smart transportation solutions. Based on technologies such as artificial intelligence, software testing, digital twins and big data, the company has been committed to the research and development and exploration of the field of autonomous driving. It has a high-quality R&D team and independently developed the autonomous driving simulation and data closed-loop tool chain SYNKROTRON® Oasis The product series can provide high-precision sensor models, dynamics models and perception-level traffic environment simulation solutions. Customers can carry out large-scale simulation testing and model training of autonomous driving systems on the simulation platform to identify defects in the autonomous driving system in advance. Reduce the cost of real vehicle testing, eliminate safety hazards in scenario implementation, and accelerate the safe implementation of autonomous driving technology in scenarios.