In 2019, the industry widely anticipated 2025 as the milestone for the large-scale deployment of Vehicle-Infrastructure Cooperation (VIC) in mass-produced vehicles. Now, past the midpoint of 2026, the number of mass-produced vehicles on the road genuinely utilizing VIC remains negligible. Some describe the VIC market as all bark and no bite, with certain projects slowing down or being suspended, and the industry as a whole cooling off from the high expectations of 2024.
On one hand, there are tens of billions of yuan in investments and an extensive list of pilot cities; on the other hand, public awareness is virtually zero. As we move through 2026, how far has VIC actually progressed? Let's delve into this topic today. If you are interested, feel free to join the discussion in the Smart Driving Frontier exchange group.
From Vehicle-Infrastructure Cooperation to Vehicle-Road-Cloud Integration?
The earliest concept of VIC referred to the information interaction between vehicles and roadside equipment, meaning the vehicle informs the road of its status, and the road informs the vehicle of traffic lights, road conditions, and other information. However, this concept has been replaced by Vehicle-Road-Cloud Integration (VRCI) in recent years.
Building upon vehicle-road interaction, VRCI adds a cloud control platform to form a three-end collaboration among vehicles, roads, and the cloud. The cloud control platform aggregates data uploaded by roadside perception devices and vehicles, processes it, and then distributes useful information to the vehicles that need it. The vehicle is responsible for execution; the road is responsible for perception; the cloud is responsible for decision-making and scheduling.
The advantage of this architecture lies in enabling vehicles to achieve beyond-line-of-sight perception capabilities, allowing them to anticipate risks ahead even through obstacles or from several kilometers away. From a technological perspective, the currently highly popular single-vehicle intelligence addresses how well a vehicle can drive itself, whereas VIC addresses how much a vehicle can know. The two are complementary rather than substitutive.
The cloud control platform is at the core of the VRCI system. It belongs to a field characterized by heavy software, strong integration, and high customization. Currently, the industry has initially formed a model of "1 cloud control basic platform + N application scenarios," which includes traffic light coordination, blind-spot warnings, and automated parking. The future technological evolution may move towards a three-tier collaborative architecture of central cloud + edge cloud + endpoint, to support the large-scale application of L4 autonomous driving.
Where Has Vehicle-Road-Cloud Integration Been Promoted?
In terms of VRCI infrastructure coverage, the deployment of infrastructure has actually progressed quite rapidly in some regions. In July 2024, the Ministry of Industry and Information Technology (MIIT) and four other departments announced the first batch of 20 national pilot cities for VRCI applications. By July 2026, this batch of pilots has entered its concluding stage. In just two years, roadside perception devices, communication units, and cloud control platforms have been built from scratch in these 20 cities.
Beijing is the city with the most vigorous promotion of VRCI. Yizhuang District has achieved functional coverage of 60 square kilometers, deployed over 600 Intelligent Connected Vehicles (ICVs), accumulated over 13 million kilometers of testing mileage, and served over 1 million trips. The Tongzhou District VRCI 3.0 project covers 175 square kilometers across the entire area and has completed intelligent upgrades for more than 500 core intersections.
In July 2026, Beijing released bidding plans for five additional districts: Miyun, Huairou, Mentougou, Daxing, and Fangshan, with an estimated investment of 835 million yuan and a construction scale of 630 square kilometers. According to incomplete statistics, Beijing centrally disclosed 8 VRCI bidding plans within 21 days, involving 16 districts, with a total investment of approximately 4.586 billion yuan.
Jiading District in Shanghai has also built the country's first IPv6+ ICV network, completed upgrades for nearly 300 smart intersections, and opened 1,117 kilometers of testing roads.
Wuhan Economic and Technological Development Zone has completed the intelligent transformation of106 kilometers of roads, opened 750 kilometers of testing roads, issued a cumulative total of 458 testing licenses, and accumulated over 3 million kilometers of testing mileage, achieving the country's first fully driverless commercial operation pilot.
In addition, traffic lights at over 1,700 intersections in Wuxi have completed intelligent upgrades; Shiyan plans to complete infrastructure transformation for 228 intersections and road sections in the urban area by the second half of 2026, basically achieving full coverage of the main urban area; and Tongxiang has become the first city in the country to achieve full-domain coverage of 5G VIC.
Judging from these data, roadside infrastructure has indeed been rolled out over the past two years, but these investments are highly concentrated in the 20 pilot cities. The "2026 China Vehicle-Road-Cloud Integration Market Research Report" released by the Savien Research Institute shows that in 2025, the investment scale in demonstration cities accounted for nearly 90%, while the number of projects and capital investment in non-demonstration cities were relatively limited.
Beijing is currently the city with the largest investment, but even in Beijing, coverage is concentrated in specific areas such as Yizhuang and Tongzhou; Wuxi took the lead nationwide in achieving networked and coordinated control for 4,253 signal-controlled intersections, but such cities remain a minority nationwide.
Furthermore, there is also an imbalance within the pilot cities themselves. Projects usually prioritize covering core road sections, while the transformation progress in suburbs and non-arterial roads is significantly lagging. Outside the first batch of 20 pilot cities, except for a few cities that actively followed up, most remain in a wait-and-see state.
Overall, the coverage of VIC is currently still point-like, and there is still a considerable distance to go before forming a network effect that ordinary car owners can continuously perceive.
Trend Changes Behind Market Data
From the perspective of the global market, the data provided by different institutions vary significantly and need to be examined separately.
Data from Global Market Insights shows that the global automotive V2X market size was estimated at $3.98 billion in 2025; Research and Markets uses a slightly different statistical caliber, valuing the global automotive V2X market at $3.9 billion in 2025; Fortune Business Insights gives a lower figure of $2.49 billion for 2025; if the statistical scope is expanded from automotive V2X to the broader V2X communication market, a report by GII (Global Industry Analysts) shows a baseline scale of $26.83 billion in 2025.
Domestic market data is relatively clear. According to estimates by Guolian Securities, the cumulative investment scale for major VIC IT equipment such as RSUs (Roadside Units), OBUs (On-Board Units), high-definition maps, and edge computing units is expected to reach 128.3 billion yuan in 2026 and 283.4 billion yuan in 2030. Intelligent roadside infrastructure is the fastest-growing segment among them, with an estimated output value increment of 22.3 billion yuan in 2025, expected to reach 417.4 billion yuan in 2030, representing a compound annual growth rate (CAGR) of 79.7%.
The "Forecast Report on the Output Value Increment of the Vehicle-Road-Cloud Integrated Intelligent Connected Vehicle Industry" released by the China Society of Automotive Engineers (China SAE) and other institutions shows that, under a neutral expectation scenario, the output value increment of the VRCI ICV industry will be 729.5 billion yuan in 2025, increasing to 2,582.5 billion yuan in 2030.
The report further breaks down that the output value increment for the ICV part will be 645.1 billion yuan in 2025 and 2,026.6 billion yuan in 2030; the output value increment for the intelligent roadside infrastructure part will be 22.3 billion yuan in 2025 and 417.4 billion yuan in 2030.
In terms of industry scale estimation, the "Forecast Report on the Output Value Increment of the Vehicle-Road-Cloud Integrated Intelligent Connected Vehicle Industry" jointly released by the China Society of Automotive Engineers, the National Intelligent Connected Vehicle Innovation Center, and other institutions shows that, under a neutral expectation scenario, the output value increment of the VRCI ICV industry will be 729.5 billion yuan in 2025, increasing to 2,582.5 billion yuan in 2030. Among them, ICVs contribute the main share, with the output value increment expected to be 645.1 billion yuan in 2025, accounting for about 88.4% of the entire industry, and 2,026.6 billion yuan in 2030, accounting for about 78.5%.
Intelligent roadside infrastructure is a relatively fast-growing segment. According to the report's forecast, the output value increment of roadside infrastructure will be 22.3 billion yuan in 2025, accounting for about 3.1% of the entire industry, and will increase to 417.4 billion yuan in 2030, with the proportion rising to 16.2%. The CAGR of the output value increment for roadside infrastructure is 79.7%, leading the growth rate among all sub-segments of the ICV industry. The installation cost of roadside equipment per kilometer has dropped from the early 4 million yuan to around 500,000 yuan, providing conditions for subsequent large-scale deployment.
In terms of the actual progress of pilot promotion, as of October 2025, the 20 VRCI application pilot cities have cumulatively opened 35,000 kilometers of testing and demonstration roads, deployed over 11,000 sets of intelligent RSUs, and cumulatively formulated and released 88 national and industry standards. A total of 17 national-level testing demonstration zones have been built nationwide, with over 10,000 testing licenses issued cumulatively. Changchun City announced a plan to invest 12.7 billion yuan over the next three years to promote VRCI construction, and related projects in multiple cities such as Beijing, Wuhan, and Ordos have also initiated bidding.
Where Exactly is Commercialization Bottlenecked?
If we only look at government planning and local construction progress, VIC seems to have entered the fast lane, but the actual market feedback is not so optimistic. Zhang Xiang, a guest professor at Huanghe Science and Technology College, told the media that achieving nationwide coverage of VRCI currently still requires massive capital investment, and the industry has not yet formed a complete commercial closed loop. The core crux lies in the fact that underlying technologies such as autonomous driving are not yet mature, and the conditions for full-scenario commercial implementation are insufficient.
In addition, the unclear operating entities and ambiguous division of rights and responsibilities are also unavoidable issues.
The operating entities are currently unclear. Whether vehicle data services should be the responsibility of automakers or platform companies, whether both parties face users jointly, or whether automakers treat it as part of their self-operated functions—there is still no unified understanding on these matters.
The division of rights and responsibilities is equally ambiguous. If roadside information errors or failures lead to safety accidents, should the responsibility be attributed to the processing link of the on-board system, the network transmission process, or the data production at the cloud or road end? Currently, there is a lack of clear laws, regulations, and accident handling rules to define this.
Cui Dongshu, Secretary-General of the Passenger Car Market Information Joint Conference of the China Automobile Dealers Association, pointed out that the industry's technological route is not yet clear, with the two routes of "road-heavy" and "vehicle-heavy" running in parallel for a long time. The equipment standards in various regions are not unified, and the resulting problem of repeated investment is also quite prominent.
From a regulatory perspective, the construction of the standard system is also catching up. In April 2026, the national standard "Technical Requirements for Information Interaction in Vehicle-Infrastructure Cooperation Part 3: Cloud Control Platform and Third-Party Application Services" was officially established; in the same month, "Technical Requirements for Information Interaction in Vehicle-Infrastructure Cooperation Part 1: Between Roadside Facilities" also entered the formulation process; in June, the China Intelligent Transportation Systems Association also approved and released 5 group standards including "Edge Computing Facilities for Vehicle-Infrastructure Cooperation Part 2: Operating System".
In addition, the international standard for VIC safety, led by the China Automotive Engineering Research Institute (CAERI), has been officially approved as a work item at ITU-T. It relies on the research results of the China SAE group standard T/CSAE 295.7—2026. There is still a distance from the formulation of the standard to its implementation, and differences in technological routes and equipment selection among various regions still exist.
The actions on the automaker side are also worth noting. FAW plans to equip a mass-produced Hongqi model with functions such as traffic management information prompts and abnormal vehicle information prompts within the year; Changan Automobile adopts a dual-track strategy of lightweight APP deployment and deep integration with the intelligent driving system, planning to deepen the development of L3-level cooperative driving in 2026; SAIC-GM-Wuling, NIO, King Long, and Zhongtong Bus, among other automakers, are all promoting the integration of these functions into vehicles. However, these functions are currently still mainly information prompts, and there is still a gap from true cooperative control.
Having discussed so much, let's revisit the initial question: how far has VIC actually progressed? Smart Driving Frontier believes that, in terms of infrastructure construction and policy promotion, it is indeed advancing, and at a considerable speed. However, in terms of the breadth and depth of commercialization, it is far from taking off. The current state is more like a massive infrastructure network being laid out, but the mechanism to truly make this network operate and generate sustainable commercial value has not yet been fully established.
Whether the experience of pilot cities can be replicated to more places, whether data services can form a stable business model, and whether automakers have enough motivation to install VIC functions on a large scale—the answers to these questions may only gradually become clear after the next batch of pilots concludes.
#AutonomousDriving #VehicleRoadCloudIntegration #VIC