Autonomous Vehicle Competitions in China

Autonomous Vehicle Competitions in China

Definition and Core Positioning
Autonomous vehicle competitions in China refer to a series of competitive events jointly organised by Chinese government agencies, industry associations, universities, and tech enterprises, focusing on autonomous driving technology validation, algorithm development, and commercial deployment. Unlike conventional motorsport, the core objective of these events is not outright speed or endurance, but rather the perception accuracy, decision-making reliability, and vehicle control stability of autonomous systems across complex driving environments. These competitions typically comprise two key segments: virtual simulation trials and on-track real-vehicle challenges. Competing entries encompass university research teams, vehicle manufacturers (OEMs), autonomous driving tech startups, and independent enthusiasts, establishing these events as a vital platform for talent development and continuous tech innovation within China’s intelligent and connected vehicle (ICV) ecosystem.

China Future Intelligent Vehicle Challenge
Inaugurated in 2009 and co-organised by the Department of Information Sciences of the National Natural Science Foundation of China and the Chinese Association of Automation, the China Future Intelligent Vehicle Challenge stands as the country’s earliest, longest-running, and most prestigious autonomous driving competition. Backed by the National Natural Science Foundation of China's Major Research Plan on "Cognitive Computing of Visual and Auditory Information", the event serves as an integrated real-world testing bed for cutting-edge autonomous driving research. The 15th edition in 2025 took place in Changshu, Jiangsu, under the theme "Advanced Autonomous Driving and Multi-Agent Embodied Interaction", attracting 14 competing teams from premier institutions including the Hong Kong University of Science and Technology (Guangzhou), Xi'an Jiaotong University, Tsinghua University, and Shanghai Jiao Tong University. This edition broke new ground by introducing humanoid and quadruped robots as active road users, requiring autonomous vehicles to recognise hand signals from robotic traffic police, interact with robotic valets, and handle extreme hazards such as falling cargo from lead vehicles, simulated heavy rain and fog, and sudden cut-ins by non-cooperative road users. The "Wukong" autonomous vehicle team from the Hong Kong University of Science and Technology (Guangzhou) clinched the top prize with the highest overall score, deploying an on-board lightweight visual recognition model capable of identifying humanoid robot gestures alongside human-like motion planning and decision-making algorithms.

i-VISTA Autonomous Vehicle Grand Challenge
Staged at the China Automotive Engineering Research Institute’s i-VISTA Intelligent Vehicle Integrated System Test Zone, the i-VISTA Autonomous Vehicle Grand Challenge is a premier fixture of the Smart China Expo, held annually in Chongqing since 2018. The competition’s technical framework is underpinned by the "China Intelligent Vehicle Index" rating system, with testing parameters modelled directly on Chinese naturalistic driving databases and local driver behavioural patterns. The event features multiple competitive categories, including the ADAS Driver Assistance Challenge, Urban Traffic Scenario Challenge, Innovative Application Challenge, Commercialisation Progress Challenge, and Low-Speed Autonomous Driving Challenge, providing tailored benchmarks across varying levels and maturity stages of autonomous technology. A distinctive highlight of the event is the inclusion of private car owners, pitting the driver assistance systems of mass-market production cars directly against dedicated autonomous setups from professional development teams. The Urban Traffic Scenario Challenge tests vehicles across more than ten real-time scenarios—spanning active safety interventions, traffic efficiency, and connected information services—under strict time limits to rigorously assess sensory precision and rapid decision-making.

CCF Intelligent Vehicle Competition - Autonomous Driving Simulation Track
Organised by the China Computer Federation, the CCF Intelligent Vehicle Competition features an Autonomous Driving Simulation track built on actual industrial case studies from Baidu Apollo, centring on autonomous driving software algorithms within urban road networks. The underlying vehicle dynamics models and digital map environments are supplied by Apollo; competitors develop and fine-tune their algorithms in a local sandbox environment before submitting their code to the Apollo Studio cloud evaluation platform. The 2026 finals were conducted online under a demerit point system, starting with a base score of 100 points per scenario with a 90-second cutoff; any collision or road boundary infraction resulted in an immediate zero score for that test run. This track is designed to hone contestants’ software development and debugging skills, serving as an incubator for practical, industry-ready engineering talent.

Baidu Apollo Intelligent Driving Special Competition and Real-Vehicle Challenge
Held in partnership with the China Robot & Artificial Intelligence Competition, Baidu Apollo has organised the Intelligent Driving Special Competition annually since 2021. Today, it ranks among China’s largest and most authoritative autonomous driving showcases, drawing over 20,000 competitors across more than 600 tertiary institutions. The 28th edition finals in 2026 were hosted in Xiangcheng, Suzhou, featuring 122 finalist teams from 120 universities nationwide and introducing a pioneering three-pillar assessment matrix: "Algorithms + Practical Execution + Commercial Viability". In addition to traditional simulation and live track events, the competition introduced a business pitching round, requiring teams to operate as tech startups and pitch commercialisation strategies to a panel of venture capital judges. Meanwhile, the Real-Vehicle Challenge brought live traffic conditions to the testing circuit, requiring teams to deploy on-board code to physical vehicles to demonstrate smooth autonomous cruising, pinpoint pedestrian avoidance, and precise execution of complex manoeuvres such as stopping accurately at pedestrian crossings. According to the head of the Baidu Apollo Open Platform, the event curriculum integrates Apollo’s live industry resources and application cases to bridge the gap between academic education and real-world automotive industry demands.

Intelligent Driving "Fengshenbang" Challenge
Hosted by the National Intelligent Connected Vehicle Quality Inspection and Testing Centre (Hunan), the Intelligent Driving "Fengshenbang" Challenge focuses squarely on real-world closed-course evaluations of series-production intelligent driving systems. The 3rd edition in 2026 introduced a dedicated category for purpose-built autonomous delivery and utility pods, featuring production models from over ten leading industry players including Neolix, White Rhino, JD Logistics, Zelos, and Youjia Innovation. The competition also debuted an authentic "High-Risk & Accident" test track that replicated actual road crash trajectories and environmental variables at a 1:1 scale, including stalled vehicles on expressways, erratic traffic flows caused by roadworks, and jaywalking pedestrians. Conducted without scripted routines or prior dry runs, vehicles relied solely on their production-spec autonomous driving suites and on-board algorithms for real-time sensing and evasive manoeuvres.

Key Technical Highlights of China's Autonomous Motorsport Ecosystem
China’s autonomous vehicle competitions demonstrate three defining technical hallmarks. First, a dual emphasis on virtual simulation and real-world testing: while the CCF IVC and Apollo Special Competition provide robust virtual simulation tracks, the Future Intelligent Vehicle Challenge and i-VISTA emphasise physical vehicle trials, creating a seamless validation pipeline from digital testing to tarmac. Second, rapidly escalating scenario complexity: evaluations have expanded from simple structured road layouts to complex unstructured environments involving multi-agent robotic interactions, adverse weather simulations, and real-world crash recreations, mirroring the industry’s push to transition autonomous driving from controlled R&D labs to complex open-road conditions. Third, an explicit focus on commercialisation: with business pitch rounds in the Apollo Special Competition and commercial pod trials in the "Fengshenbang" Challenge, assessments now reward practical market viability alongside pure algorithmic performance, covering the entire lifecycle from early R&D to mass-market production readiness.