news

From August 16 to 17, 2024, the 4th Shenyang Intelligent Connected Vehicle Conference with the theme of "Vehicle-Road Collaboration, New Quality Empowerment" will be officially held in Dadong District, Shenyang. The conference is co-hosted by the Shenyang Municipal People's Government and the China Electric Vehicle Hundred People's Forum. Leaders of national ministries and commissions, relevant leaders of local governments, well-known academicians and experts, domestic and foreign corporate representatives, and mainstream media and other elites from all walks of life are invited to discuss key issues such as the development trend, commercialization exploration, and operation model of the intelligent connected vehicle industry, and work together to promote the high-quality development of the intelligent connected vehicle industry.

As one of the core topics of the conference, the large-scale development and application of vehicle-road-cloud integration has received great attention and heated discussions from the government, industry and enterprises. Guo Xingrong, CTO of Mushroom Car Union, attended the Forum on Application Practice and Business Exploration of Vehicle-Road-Cloud Integration and fully demonstrated to the industry Mushroom Car Union's mature practices and technological innovations in the field of vehicle-road-cloud integration.

Guo Xingrong said that when the construction of vehicle-road-cloud integration will move from the small-scale testing and verification stage to the large-scale implementation stage, how to make roadside infrastructure play a greater value is the key to ensuring the effectiveness of vehicle-road-cloud integration. In this process, technical breakthroughs and innovations are needed in four aspects: 1) whether the software and hardware have formed standardized products, 2) whether the three ends of vehicle, road and cloud are connected and linked, 3) how to generate usable and useful high-quality data, and 4) how data can enable applications. Among them, data is both a basic resource and a strategic resource, and it is also the core driving force for the sustainable development of vehicle-road-cloud integration.

In December 2023, the National Data Bureau and 17 other departments recently issued the "Three-Year Action Plan for "Data Element ×" (2024-2026)", proposing the implementation of the "Data Element × Transportation" action, including exploring the value of data reuse, promoting the innovative development of intelligent connected vehicles, and other aspects, to accelerate the construction of smart transportation.

As a new production factor, data is the foundation for the digitalization and intelligence of the transportation industry. Traffic data elements include various types of data such as traffic flow, vehicle conditions, driving behavior, and traffic facilities, which are characterized by large scale, diverse types, and low value density. Under the vehicle-road-cloud integrated system, high-quality and highly available data can promote traffic flow management, improve the efficiency of autonomous driving decision-making, and enhance the accuracy of traffic situation analysis and prediction, which plays an important role in improving the safety of autonomous driving and road traffic capacity.

Guo Xingrong further explained that highly available data means that the data can truly reflect the road traffic conditions and ensure that the data can be called in a timely and accurate manner when the user needs it, ensuring the effectiveness of the system decision-making. High-quality data has the characteristics of accuracy, completeness, consistency, timeliness and reliability, ensuring that traffic data can be trusted and used. If the data quality is poor, the infrastructure will not be able to function and may lead to wrong decisions or operations, which is unacceptable for the transportation industry, which is based on safety. Therefore, only by generating high-quality and highly available data can an accurate basis be provided for the collaborative analysis and global decision-making of vehicle-road-cloud integration.

How should we achieve high availability and high quality of traffic data for vehicle-road-cloud integration? Guo Xingrong introduced that the premise of high data availability is the free flow and full sharing of data. A major challenge in the current construction of vehicle-road-cloud integration is that the data of the three ends of vehicle, road and cloud are separated from each other. A lot of data is deposited in the respective systems of relevant departments and automobile companies. There are a large number of "data chimneys" and "information islands". The cross-domain flow and sharing of data are not realized, which limits the value of data.

In response to this, Mushroom Autolink has created a complete solution for the linkage of vehicle, road and cloud, successfully breaking down the data barriers between road, cloud data and intelligent connected vehicles, and building an integrated vehicle-road-cloud system with the same architecture, unified standards, security and reliability for governments and intelligent connected data operation service providers, paving the way for the collaborative processing and application of traffic data.

On the roadside, Mushroom AutoLink independently develops AI digital road base stations and roadside perception systems (MRS) and other software and hardware products, which can perceive road traffic elements in real time, all-weather and non-stop, and form real-time digital twins of intersections and continuous sections based on traffic big data, providing beyond-visual-range perception capabilities for intelligent connected vehicles. The roadside perception system (MRS) is embedded with Mushroom AutoLink's BigFusion model and VAE prediction model, integrating data from multiple types of sensors such as lidar, millimeter-wave radar, and cameras. After deep neural network processing, it completes tasks including panoramic segmentation, object detection, scene flow estimation, motion state estimation, self-vehicle posture correction, and occupied voxel estimation in three-dimensional scenes, and forms the motion trajectory of moving objects in the next 5 seconds, the vehicle's driving intention, the self-vehicle's decision information, and the self-vehicle's future planning trajectory information, helping intelligent connected vehicles to analyze traffic situations and provide global optimal path planning.

Recently, after testing and issuing a report by China Automotive (Beijing) Intelligent Connected Vehicle Research Institute Co., Ltd., Mushroom Autolink AI digital road base station and MRS system have reached the highest requirements of "double SL3". The system perception time accuracy of the AI ​​digital road base station reached 4.4 milliseconds, and the perception delay was 41.57 milliseconds, both exceeding the SL3 standard. At the same time, the positioning accuracy, heading angle and size detection accuracy have reached the industry-leading level.

In terms of cloud, Mushroom Autolink has built a smart transportation AI cloud platform, which can be regarded as the "super brain" of urban traffic. It can intelligently identify, analyze and integrate massive heterogeneous traffic data such as images and videos, transform static data into dynamic knowledge and traffic portraits, and achieve accurate judgment of traffic development trends. At the same time, relying on artificial intelligence engines and algorithm computing engines, it can achieve collaborative decision-making and control between cars, cars and roads, and cars and clouds, and can perform efficient calculations and analysis on specific traffic events, further accurately predict traffic flow, dynamically adjust signal light timing, effectively manage and optimize traffic flow, and provide strategic guidance for improving the traffic capacity of urban roads.

For autonomous vehicles, Mogo AutoLink has developed a complete MogoAP autonomous driving system, which is perfectly connected with the roadside MRS system and can be installed on autonomous buses, street sweepers, patrol cars, delivery vehicles and other types of autonomous vehicles. For example, the first pre-installed mass-produced autonomous driving bus developed by Mogo AutoLink, in addition to standard V2X functions and L4 autonomous driving capabilities, is also the first to realize the application of semi-solid laser radar and self-developed Orin domain controller on bus models, helping to upgrade the smartness of urban public transportation.

At the same time, Mushroom Car Alliance uses technical means such as multi-source and multi-terminal data fusion, multi-modal information fusion, ultra-low latency transmission network, collaborative perception decision-making and control to deeply integrate intelligent connected vehicles, roadside infrastructure and cloud platforms. From a global perspective, through multi-vehicle collaboration, intention sharing, collaborative decision-making and other methods, the system comprehensively improves the accuracy of overall traffic prediction and decision-making efficiency, and finds the "optimal solution" to alleviate traffic congestion and improve traffic safety.

Based on the above technological innovations, Mushroom Car Union has formed a set of mature solutions that can be implemented, replicated and promoted, and has been deployed in many places including Beijing, Shanghai, Shenzhen, Tianjin, Sichuan, Liaoning, Hunan, Yunnan, Shandong, Hubei, etc., covering various scenarios such as public transportation, scenic spots, and park transfers, further verifying the implementation capabilities and deployment results of its solutions.

With the in-depth development of national-level Internet of Vehicles pilot zones, smart city infrastructure and intelligent connected vehicles collaborative development pilot projects, intelligent connected vehicle access and road access pilot projects, and intelligent connected vehicle "vehicle-road-cloud integration" application pilot projects, vehicle-road-cloud integration is bursting with strong vitality. Driven by the power of scientific and technological innovation, vehicle-road-cloud integrated autonomous driving companies represented by Mushroom Auto will help more cities move towards intelligent upgrades and contribute to the grand blueprint of vehicle-road-cloud integration.