2023 "CCF Scientific and Technological Achievement Award" selection results announcement

The CCF Award Committee decided to award the 2023 CCF Scientific and Technological Achievement Award to 41 projects, including the project of "Basic Theory and Method of Data Mining for Recommendation System", the project of "Key Technologies and Applications of Virtual and Real Fusion Spatial Computing for Complex Equipment Support", and the project of "OpenHarmony Intelligent Terminal Operating System Base".

The "CCF Scientific and Technological Achievement Award" is awarded to outstanding achievements with important discoveries, inventions and original innovations in the field of computer science, technology or engineering, and have a certain international impact in related fields.

This year, a total of 95 recommended/declared projects were received, and after selection, the CCF Award Committee decided to award the 2023 CCF Scientific and Technological Achievement Award for 41 projects, including the project of "Basic Theory and Method of Data Mining for Recommendation System", the project of "Key Technologies and Applications of Virtual and Real Fusion Spatial Computing for Complex Equipment Support", and the project of "OpenHarmony Intelligent Terminal Operating System Base".

This is hereby announced.

Chinese Computer Society

September 27, 2023

Attached: List of winning projects of the 2023 "CCF Scientific and Technological Achievement Award"

First Prize in Natural Science (1 award)

Project Title: Basic Theory and Method of Data Mining for Recommender System

Completed by: University of Science and Technology of China, Microsoft Research Asia

Main finishers: Chen Enhong, Lian Defu, Xie Xing, Lian Jianxun

Aiming at the technical challenges faced by the recommendation system in terms of hidden and difficult to understand, sparse and difficult to use, and complex and difficult to match and efficient, this project has achieved key theoretical and method breakthroughs with "interest reasoning fusion-recommendation model generalization-retrieval reasoning extension" as the main line, comprehensively improved the generalization performance and reasoning efficiency of the recommendation algorithm, and the paper won the ACM SIGKDD Best Student Paper Award, and the relevant recommendation system was open-sourced, and the results were successfully applied in scenarios such as advertising, product recommendation, and news recommendation.

Second Prize in Natural Science (5 awards)

Project Title: Theory and System of Iterative Computation of Large-scale Graphs in Cloud Environment

Completed by: Northeastern University

Main finishers: Yu Ge, Zhang Yanfeng, Gu Yu, Wang Zhigang, Gong Shufeng, Wang Qiange

This project focuses on the theory and methods of large-scale graph iterative computation in cloud environment, constructs the basic theory and method system of high-time-sensitive large-scale graph iterative computation from three aspects: parallel model, iterative execution method and scalability, reveals the monotonicity characteristics and convergence contribution difference properties in large-graph iterative computing, and the relevant results won the ACM SOCC Outstanding Paper Award, and the relevant theories and methods have been applied in enterprises such as China Mobile and Alibaba.

Project Title: Heterogeneous Wireless Network Coexistence and Convergence Theory and Method

Completed by: Tsinghua University, Beijing University of Posts and Telecommunications

Main finishers: He Yuan, Guo Xiuzhen, Zheng Xiaolong, Liu Yunhao, Wang Jiliang

Focusing on the problem of coexistence and convergence of heterogeneous wireless networks, this project proposes the principle of high-reliability mutual identification and coexistence, the principle of high-availability side-channel communication and the high-efficiency physical layer fusion method, which solves the problems of information isolation between heterogeneous terminals, lack of cross-domain information exchange media and mismatch of converged communication capabilities, and the relevant results are applied in the online monitoring of the condenser camera of the State Grid UHV converter station, the online monitoring of the production environment of Jiangsu Nangang, and the integrated intelligent platform of Xi'an Happy Forest Belt Transport Management and Control.

Project Title: Basic Theory and Method of Large Graph Data Condensation Subgraph Mining

Completed by: Beijing Institute of Technology, Hong Kong Chinese University, Shenzhen University

Main finishers: Li Ronghua, Yu Xu, Yuan Ye, Mao Rui, Wang Guoren

This project establishes a set of basic theories and methods for large-graph data condensation subgraph mining, including the influence modeling method, asymptotic calculation theory and incremental maintenance theory of condensed subgraph of large-graph data, and realizes the technical breakthrough of large-graph data condensation subgraph mining from "difficult to calculate" to "computable" to "second-level response", and the results won the Best Paper Award and the Best Student Paper Award of ICDE, and have been successfully applied in CSIC, ByteDance and other units, and achieved good social and economic benefits.

Project Title: Network Architecture and Resource Management Theory and Methods for Software-Defined Cloud Data Centers

Completed units: National University of Defense Technology, Academy of Military Sciences

Main finishers: Guo Deke, Zhu Xiaomin, Luo Lailong, Ren Bangbao, Wang Ji, Xie Junjie

Aiming at the basic problems of software-defined cloud data centers, this project puts forward theories and methods such as scalable network architecture design, multi-dimensional resource collaborative management, cloud computing task collaboration and fault-tolerant scheduling of software-defined cloud data centers, which improves the resource utilization efficiency and task service quality of cloud data centers, and the results are applied in Internet enterprises and open source platforms such as Alibaba Cloud and Hadoop.

Project Title: Theory and Methods of Multichannel Simulation of Natural Phenomena

Completed units: Tianjin University, Zhejiang University

Main finishers: Liu Shiguang, Wang Zhangye, Peng Qunsheng

Focusing on the multi-channel co-simulation of natural phenomena, this project proposes a series of new methods such as two-phase flow model, modeling and mapping of tornadoes and sandstorms, and liquid-liquid mixed flow simulation, which realizes the synchronous synthesis of multi-channel information such as audiovisual and haptic sensation, and the relevant results are applied in the virtual simulation of the national nuclear power test process, the experimental measurement and control simulation of the Nuclear Power Research and Design Institute, and the film and television production.

First Prize for Technological Invention (3 Awards)

Project Title: Key Technologies and Applications of Virtual and Real Fusion Spatial Computing for Complex Equipment Support

Completed by: Beihang University, Beijing Jinfa Science and Technology Co., Ltd

Main finishers: Wang Lili, Zhou Dong, Zhou Bin, Zhao Qichao, Lei Xiaoyong, Zhao Qinping

This project focuses on the requirements of virtual and real space presentation, interaction and maintainability evaluation for complex equipment support, breaks through a series of key technologies of virtual and real fusion space computing, builds a virtual and real fusion support platform for complex equipment support, and develops a virtual and real fusion support application system for domestic strategic equipment in the aerospace field such as on-orbit space stations, large civil aviation passenger aircraft C919 and C929, a certain turboshaft aircraft engine, and large UAV Changying, and achieves significant social and economic benefits.

Project Name: Key Technologies and Applications of Cloud-Native Runtime Systems for Internet Services

Completed units: Shanghai Jiao Tong University, Alibaba Cloud Computing Co., Ltd., Lenovo (Beijing) Co., Ltd

Main finishers: Chen Quan, Allergic Yi, Ma Tao, Song Zhuo, Zhang Wenjie, Yang Yong

In response to the demand for cloud native for Internet services, the project breaks through key technologies such as dataflow-driven workflow efficient scheduling, multi-level shared container runtime and startup optimization, performance conflict elimination of hybrid deployment containers, and QoS rapid recovery of dynamic microservice graphs, and builds a cloud-native runtime system "Xiaoyao", integrated in Alibaba Cloud Linux and Lenovo xCloud cloud-native container cloud platform, serving a number of industry-leading customers. Achievements: 6 open source software.

Project Title: Key technologies and applications of service collaborative computing for distributed AI applications

Completed by: Beijing University of Posts and Telecommunications, Inspur (Beijing) Electronic Information Industry Co., Ltd

Main finishers: Wang Shangguang, Zhou Ao, Wang Yanwei, Sang Xinzhu, E Haihong, Du Junping

This project has made a series of technical inventions and innovations in fast and efficient service collaboration, green and reliable computing resource scheduling, and computing acceleration of heterogeneous chip full interconnection, alleviating the operational inefficiency problems faced by distributed AI applications caused by limited computing power, unbalanced resource scheduling, and insufficient service supply, and the results have been applied to the three major operators, large Internet companies and other units, and achieved good economic and social benefits.

Second Prize for Technological Invention (3 Awards)

Project Title: Key technologies and systems for integrated large-scale cloud network operation and maintenance with digital intelligence integration

Completed by: Tsinghua University, Alibaba Cloud Computing Co., Ltd., Zhongguancun Laboratory

Main finishers: Yang Jiahai, Zhu Shunmin, Dong Enhuan, Lv Biao, Wang Zhiliang, Zhang Shize

Aiming at the cost, efficiency, and real-time challenges faced by cloud network O&M, this project breaks through key technologies such as high-performance O&M data management, unattended network change, high-precision network monitoring, and full-link abnormal self-healing, and develops an integrated large-scale cloud network O&M system with digital intelligence integration, improves the reliability of cloud network systems, and supports more than 3 million Alibaba Cloud customers worldwide.

Project Title: Integrated modeling and customized service technology and application of cyber-physical fusion system

Completed by: Northwestern Polytechnical University

Main finishers: Yang Gang, Dong Yunwei, Yao Yuan, Shen Bo, Zhou Xingshe

Aiming at the technical challenges of "integrated construction of fusion model, comprehensive verification of system characteristics, and broad-spectrum adaptation of operating platform" that urgently need to be solved in cyber-physical fusion system (CPS), this project proposes scenario-driven integrated modeling methods, formal verification methods for key features, adaptive operation service technology with complementary capabilities, and software-defined domain optimization technology, etc., and the results have achieved significant application results in the development of national key models, key infrastructure operation and maintenance, and the transformation and upgrading of many key enterprises.

Project Title: High-level autonomous driving predictive decision-making technology based on data closed-loop engine

Completed units: Shanghai Didi Woya Technology Co., Ltd., Beijing Didi Unlimited Development Co., Ltd

Main finishers: Zhang Bo, Sheng Kehua, Meng Xing, Liu Chunyang, Wu Guobin, Li Qun

This project combines deep neural networks and driving data to propose accurate and efficient behavior prediction and safe and intelligent vehicle decision-making algorithms, which provides important support for Didi's autonomous driving public road test, and has been applied to nearly 200 high-level autonomous vehicles deployed by Didi in Beijing, Shanghai, Guangzhou and other cities.

Special Prize for Scientific and Technological Progress (1 award)

Project name: OpenHarmony intelligent terminal operating system dock

Completed by: Huawei Technologies Co., Ltd., Shanghai Jiao Tong University, Institute of Software, Chinese Academy of Sciences

Main finishers: Chen Haibo, Liu Xiaojian, Jia Ning, Ren Gelin, Wu Yanjun, Ren Xudong

This project focuses on architecture-level innovation in the core areas of intelligent terminal operating systems such as storage, memory, scheduling, and distributed basic mechanisms, and achieves industry-leading results in a number of key indicators. The results have been widely used in Huawei's 330 million terminal products and terminals in key national industries (finance, transportation, education, Anping, energy, aerospace, etc.), and some technologies have become industry standards, contributing to the open source community and talent training of the mainland operating system.

First Prize for Scientific and Technological Progress (5)

Project Name: Key technologies and systems of ultra-large-scale cloud computing based on Dongfeng-Peugeot-Citroen software and hardware collaborative virtualization

Completed by: Alibaba Cloud Computing Co., Ltd., Shanghai Jiao Tong University

Main finishers: Zhang Xiantao, Jiang Linquan, Yao Jianguo, Zhang Jianfeng, Yang Hang, Jiang Jiangwei

Focusing on the key technologies of ultra-large-scale cloud computing, this project proposes the technical system and structure of "cloud-native virtualization" for large-scale public clouds, and makes technological breakthroughs in high-isolation and low-loss virtualization of computing and I/O separation, acceleration of software-hardware collaborative I/O virtualization, RAS and load rate improvement of software-hardware collaborative design virtualization, etc., reducing the cost of cloud computing virtualization, improving the performance and efficiency of cloud services, and the project results are applied to Alibaba Cloud and serving more than 4 million users around the world.

Project Title: Science in Science Fiction Movies: The <流浪地球2>Hand-Drawn Scientist Daddy

Completed units: Institute of Computing Technology, Chinese Academy of Sciences, China Science and Technology Press Co., Ltd., Southern University of Science and Technology

Main finishers: Wang Yuanzhuo, Cui Yuanhao, Lu Yuan, Zheng Hongwei, Shen Yinghan, Jiang Xuhui

Focusing on the popularization and dissemination of science fiction films and scientific and technological knowledge, this project creatively integrates film and television production and science communication, and publishes the popular science book "Science in Science Fiction Movies: Hand-drawn by Scientist's Daddy"<流浪地球2>, exploring a new model of popularization of computational science. The book has been fully affirmed by all walks of life, and has been continuously reported by CCTV, People's Daily and so on, achieving a one-time dissemination of 660 million popular science content and a secondary dissemination of more than 6 billion, which has a huge social impact.

Project Name: Research on domestic high-end server system design technology

Completed units: Inspur (Shandong) Computer Technology Co., Ltd., Inspur Electronic Information Industry Co., Ltd., Suzhou Inspur Intelligent Technology Co., Ltd

Main finishers: Gong Wei, Li Yan, Lv Dongbo, Xu Siqiang, Li Wenfang, Wang Yong

Aiming at the key technical bottlenecks of domestic high-end servers, this project proposes a dynamic reconfigurable server system architecture and independent controllable fault diagnosis system based on domestic processors, which fills the gap of domestic power supply solutions, improves system stability and reliability, improves the overall autonomy of the mainland server industry chain, and the results have been widely used in key fields such as party and government, communications, finance, energy, and the Internet.

Project Name: Key technologies and system platforms for device-cloud collaboration for diversified intelligent industrial applications

Completed by: Taobao (China) Software Co., Ltd., Shanghai Jiao Tong University, Zhejiang University

Main finishers: Lv Chengfei, Niu Chao, Zhang Shengyu, Chen Zhiwen, Wu Fan, Wu Fei

Aiming at the challenges of diversified intelligent industrial application scenarios on mobile terminals, this project developed Walle, a universal system platform for device-cloud collaborative learning, and built a new system of personalized information flow recommendation, multimodal content understanding, three-dimensional reconstruction and real-time rendering technology with device-cloud collaborative computing architecture as the core, achieved significant economic benefits in Alibaba Group's typical application scenarios, and open-sourced the high-performance lightweight deep learning engine MNN.

Project Title: Key Technologies and Applications of Accurate Guidance of Digital and Real Integration

Completed by: Lenovo Research Institute, Xi'an Jiaotong University

Main finishers: Wang Qianying, Tian Feng, Wu Yaqiang, Zhang Lingling, Wang Luyan, Zhang Xiaoping

Focusing on the key technologies of accurate guidance based on the integration of data and reality, this project proposes a series of new technologies such as knowledge system integration construction and automatic annotation, virtual and real integration interaction and intelligent question answering, paper and pen error detection and adaptive test question generation, and the Lenovo smart education tablet, education large screen and Lenovo precision guidance platform developed provide technical support for intelligent teaching in primary and secondary schools across the country. The results have been applied on a large scale in more than 10,000 schools in 31 provinces, municipalities and autonomous regions across the country, and have achieved good social and economic benefits.

Second Prize for Scientific and Technological Progress (10)

Project Name: Key technologies and applications of high-end diesel engine fault diagnosis and intelligent services

Completed unit: Weichai Power Co., Ltd

Main finishers: Tang Bo, Song Yedong, Yin Lihang, Wang Helin, Wang Yonglai, Zhang Haiwei

This project is oriented to high-end diesel engine fault diagnosis and intelligent services, and has made important progress in the data collection, calculation and application verification system of engine fault diagnosis, intelligent maintenance reminder technology, engine performance monitoring index early warning technology, intelligent guided self-learning fault diagnosis technology, etc., and the products have been successfully applied to mainstream enterprises such as Sinotruk, FAW, Shantui, etc., and fault diagnosis tools and technologies have been promoted to more than 7,000 service stations, greatly improving customer satisfaction.

Project Title: Urban Comprehensive Risk Monitoring and Early Warning System Based on Air-Space Intelligent Perception

Completed unit: Aerospace Hongtu Information Technology Co., Ltd

Main finishers: Yang Guangping, Liao Tongkui, Chen Yuanyuan, Wang Tao, Zhu Liya, Fan Lei

Aiming at the characteristics and business needs of urban natural disasters, this project has made important progress in intelligent perception based on multi-source data of space, space and ground, early warning push based on trajectory mining, auxiliary decision-making based on knowledge graph, comprehensive disaster risk monitoring and judgment and early warning decision-making, etc., which significantly enhances the capabilities of urban disaster monitoring and early warning, disaster prevention and mitigation, and emergency command and rescue, improves the service and release time of major disaster early warning information, and has been applied on a large scale in the field of emergency management of natural disasters.

Project Title: Key Technologies and Systems of Distributed Knowledge Graph Data Management

Completed by: Tianjin University, Chinese Minmin University, Tianjin Nanda General Data Technology Co., Ltd

Main finishers: Wang Xin, Chai Yunpeng, Zhao Wei

Aiming at this important field of database management system, this project breaks through key technologies such as unified model and distributed management of knowledge graph data, and the developed system supports the efficient management of knowledge graph with more than 300 million vertices and more than 1 billion edges, and the results are applied to the domestic database Nanda General GBase UP, which has achieved large-scale market application in key industries such as finance and government affairs, and achieved significant economic and social benefits.

Project Title: Intelligent deployment and acceleration of key technologies and applications of large-scale online services under cloud-network-edge convergence architecture

Completed by: Fudan University, Wangsu Technology Co., Ltd

Main finishers: Lv Zhizhi, Wang Xin, Wu Jie, Zhao Jin, Cai Longshi, Cao Jianjia

Focusing on the construction of independent property rights computing power network, this project proposes intelligent resource deployment technology, task scheduling technology, network update optimization technology and acceleration scheme for global heterogeneous service system under cloud-network-edge convergence architecture, builds a large-scale online service platform with cloud-network-edge convergence architecture, and provides cloud-network-edge convergence services for more than 200 customers' Internet portal platforms and industrial Internet applications, and achieves good economic and social benefits.

Project Title: R&D and large-scale application of ultra-large-scale financial intelligent decision-making system

Completed by: Zhejiang University, Ant Group

Main finishers: Zheng Xiaolin, Zhou Jun, Chen Chaochao, Lu Xingyu, Zhu Mengying, Wu Hua

Aiming at the problem of financial intelligent decision-making modeling in inclusive finance, this project proposes complex heterogeneous financial data insight analysis technology, ultra-large-scale high-performance operation research optimization decision-making technology and industrial-grade full-link financial decision-making intelligent service technology, develops an ultra-large-scale financial intelligent decision-making platform that supports tens of billions of parameters, and applies the results in core financial scenarios such as intelligent risk control, intelligent marketing, and wealth management.

Project Title: R&D and industry-scale application of 5G private network based on cloud native and intelligent endogenous

Completed by: AsiaInfo Technology (China) Co., Ltd., Tsinghua University, Guodian Power Shuangwei Inner Mongolia Shanghai Temple Energy Co., Ltd

Main finishers: Ouyang Ye, Zhang Yaqin, Zhu Duozhi, Ye Xiaozhou, Liu Yunxin, Peng Zhibin

This project develops a 5G private network based on cloud native and endogenous intelligence, breaks through the core technologies such as full-stack cloudification, virtualization and end-to-end slicing of the private network that takes into account both openness and performance, realizes the transformation of 5G radio access network and core network external intelligence to intelligent endogenous based on intelligent network elements, provides an effective way for the software industry to open up the software industry for the research and development of communication equipment with closed software and hardware, and has been applied on a large scale in key energy industries such as nuclear power, thermal power and wind power.

Project Name: Research and large-scale application of key technologies of AI + Security: Intelligent Security Fusion

Completed units: NSFOCUS Technology Group Co., Ltd., Nanjing University of Aeronautics and Astronautics, Shenzhen University

Main finishers: Gu Dujuan, Xue Mingfu, Wang Xingkai, Jiang Kui, Liu Wenmao, Li Wenjin

Aiming at the technical difficulties in the "AI+Security" intelligent security fusion system, this project has made important progress in all-weather and all-round monitoring of advanced persistent threats in the fusion security graph, adversarial attack detection and defense for AI models, and builds a one-stop full-process intelligent security analysis service based on XOps, and the developed products have been applied on a large scale in government, operators, finance and other industries, and achieved outstanding economic and social benefits.

Project Title: Key technologies of spatiotemporal data prediction and their applications in transportation and logistics

Completed unit: Beijing Jiaotong University, China Merchants Xinzhi Technology Co., Ltd

Main finishers: Wan Huaiyu, Guo Shengnan, Teng Zhiwei, Zhang Shuo, Wang Zhenhua, Lin Youfang

Aiming at the spatiotemporal data prediction needs in the field of intelligent transportation and intelligent logistics, this project proposes a series of effective prediction models for different types of spatiotemporal data such as spatiotemporal map data, spatiotemporal grid data and spatiotemporal trajectory data, constructs a technical system for spatiotemporal data prediction, and is widely used in the fields of highway, civil aviation and logistics, improves the operation efficiency and service quality of transportation and logistics systems, and achieves significant socio-economic benefits.

Project Name: Safe Compensation - 2-Day Quick Claim Technology and Application Based on Multimodal Artificial Intelligence

Completed by: Ant Group

Main finishers: Chen Jingdong, Wang Hongbin, Han Dong, Li Yadong, Huang Zuming, Xie Lele

Aiming at the pain points of difficult and slow claim settlement in the insurance industry, this project breaks through key technologies such as high-precision identification of complex medical voucher information, claims decision-making constrained by knowledge in the insurance field, accurate investigation of privacy computing cases, and intelligent document camera, and develops a "2-day" fast claims system with fast application, fast review and fast investigation, which greatly improves the efficiency of health insurance claims, and has been successfully applied in 93 products of 17 insurance companies, achieving significant social and economic benefits.

Project Title: SMTX OS: Hyperconverged System with Full-Link I/O Performance Optimization

Completed unit: Beijing Zhiling Haina Technology Co., Ltd

Main finishers: Zhang Kai, Zhou Yutian, Ke Jiewei, Feng Li, Wang Sai, Xu Wenhao

This project addresses the challenges of cloud computing platform infrastructure software, and makes important progress in non-volatile memory-based I/O performance optimization, RDMA-based network communication optimization, and zero copy of virtual hard disk data. The SMTX OS 5.0 series hyper-converged system developed by the project has been widely deployed and applied in finance, medical, advanced manufacturing and other industries, and has achieved significant economic and conference benefits.

Third Prize for Scientific and Technological Progress (13)

Project Name: Key Technologies and Applications of Intelligent Data Security Protection

Completed by: Beijing Yisaitong Science and Technology Development Co., Ltd., Beijing Information Science and Technology University

Main finishers: Zhu Hejun, Yi Junkai, Cui Peisheng, Liang Jinqian, Song Chunling, Xia Kun

This project focuses on breakthroughs in key technologies such as adaptive protection of endpoint data, audit of massive data content, depth perception of important data, and intelligent protection of sensitive data, and the results have been applied to finance, manufacturing and other industries, with more than 10,000 users and more than 8 million deployed terminals, protecting the security of national data assets and achieving significant social benefits.

Project Name: Electric Power Marketing Customer Service Business Middle Office and Its Key Technology Application

Completed unit: Beijing Zhongdian Puhua Co., Ltd

Main finishers: Ouyang Hong, Zhu Pingfei, Jiang Zaiyu, Xiong Genxin, Liu Junyan, Fang Hongwang

This project applies advanced methods and technologies such as unified customer model, domain-driven design, microservice architecture, and cloud computing to develop a middle platform system for power marketing customer service business, which supports the digital service business of State Grid Corporation for 470 million power customers, and achieves outstanding social and economic benefits.

Project Name: Key technologies and industrial applications of carrier-grade edge cache storage

Completed units: ZTE Co., Ltd., Institute of Computing Technology, Chinese Academy of Sciences, Tianjin University of Technology

Main finishers: Yang Hongzhang, Chen Shimin, Jiang Dejun, Tu Zhengfeng, Xu Guangping, Guo Bin

This project breaks through the key technologies such as high performance, high reliability, and easy operation and maintenance of edge cache storage, and develops ZTE distributed cache DCACHE with independent intellectual property rights, which is widely used in telecommunications, finance, government affairs, education and other fields, serving more than 100 units in more than 30 countries, and achieving significant economic and social benefits.

Project Title: Automatic replenishment system based on AI sales forecasting

Completed unit: Multipoint Life (China) Network Technology Co., Ltd

Main finishers: Xu Bin, Wu Yandi, Li Luying, Guo Jie, Luo Shili, Jiang Xiyue

Focusing on the problem of digitalization of the retail supply chain, this project developed an intelligent adaptation multi-format technology based on advanced AI sales forecasting technology, and built a scenario-based automatic replenishment system - multi-point DMALL, which has been successfully applied in multiple merchants. The project can empower retailers from the operational and management levels, and achieve obvious economic and social benefits.

Project Name: Full-stack Autonomous and Controllable Enterprise Resource Planning and Production and Operation System Based on Artificial Intelligence (Power Grid Management Platform)

Completed by: China Southern Power Grid Co., Ltd., China Southern Power Grid Digital Platform Technology (Guangdong) Co., Ltd

Main finishers: Zhao Ming, Chen Bin, Wu Zhenrong, Lin Kequan, Li Yasong, Wang Zhongjun

Facing the relevant needs of the construction of new power system, digital grid and modern power supply service system, this project proposes a multi-resource pool hybrid scheduling cloud platform technology based on storage and computing separation and one cloud and multiple cores, which realizes the large-scale application of clusters in the domestic database field, and launches a full-stack independent and controllable ultra-large ERP system, which has achieved good application results.

Project Title: Research and application of key technologies for full-link continuous delivery and energy saving of 5G network elements based on computing power network

Completed by: ZTE Corporation Limited

Main finishers: Ju Weigang, Wang Jia, Hu Jidong, Zhang Yunlong, Wang Xiangyu, Zhang Youwen

Focusing on the continuous delivery of 5G network elements and the challenges of energy conservation and consumption reduction, this project proposes key technologies for full-link continuous delivery and energy conservation based on computing power networks, realizes layered orchestration of environmental resources, parallel automatic deployment and testing, grayscale acceptance and release, and intelligent energy saving, and completes continuous delivery between equipment vendors and operators. This achievement provides strong support for 5G construction, and has achieved significant social and economic benefits in digital transformation, efficiency and cost reduction, and dual carbon energy saving.

Project Name: Key Technologies and Applications of AI Visual Quality Inspection Platform Based on 5G+MEC

Completed units: China University of Petroleum (East China), Kaos Industrial Intelligence Research Institute (Qingdao) Co., Ltd., Kaos IoT Technology Co., Ltd

Main finishers: Liu Weifeng, Lu Xiaoping, Sheng Guojun, Liu Baodi, Qin Chenggang

Aiming at the problem of quality inspection in the manufacturing industry, this project builds an architecture for computing and network integration through the integration of edge computing and 5G technology, designs visual algorithm models and edge visual inspection algorithms, and develops an AI visual quality inspection platform based on 5G + MEC, which is applied to home appliances, automobiles, electronics and other industries, and achieves outstanding economic and social benefits.

Project Title: AI-driven cloud-native intelligent scheduling technology and application

Completed by: Tencent Cloud Computing (Beijing) Co., Ltd., Sun Yat-sen University, Tencent Technology (Shenzhen) Co., Ltd

Main finishers: Zou Hui, Chen Pengfei, Yu Guangyou, Meng Fanjie, Huang Tao, Fang Rui

Aiming at the problem of cloud-native intelligent scheduling, this project proposes a long-term and short-term cloud-native resource prediction algorithm based on deep Transformer, a service scheduling algorithm based on mixed integer programming and online learning, and an active learning algorithm based on semi-supervised variational autoencoder (VAE) for service interference detection and stability assurance. Related technologies are open source and successfully applied, achieving outstanding economic and social benefits.

Project Title: Logistics AI System Based on Geographic Multimodal and Causal Inference

Completed by: Zhejiang Cainiao Supply Chain Management Co., Ltd

Main finishers: Wu Lixia, Hu Haoyuan, Li Peng, Fu Lei, Lou Junhong, Deng Minhui

Aiming at the problem of operation decision-making of complex logistics giant systems, this project develops a logistics AI system based on geographic multimodal and causal learning, proposes a multimodal dual-system technology of knowledge graph + pre-training model in the logistics field, designs a decision-making framework for causal structure learning and inference, and builds an open source ecosystem of logistics AI. The results have been applied in Cainiao's internal and external scenarios and achieved significant economic and social benefits.

Project Name: Key Technologies and Applications of Wide Area Location Services

Completed units: Tongji University, Qianxun Location Network Co., Ltd., China Mobile Intelligent Network Technology Co., Ltd

Main finishers: Liu Eryu, Chen Jinpei, Feng Shaojun, Jiang Xin, Rao Weixiong, Qu Xinyu

Focusing on the problem of building a wide-area location service system, this project proposes a "cloud integration" cross-network convergence service architecture, breaks through the small-sample learning and trajectory recovery technology of cellular positioning, and the indoor Beidou positioning technology of the same standard as indoor and outdoor, and builds the world's largest 5G+ Beidou high-precision positioning system and location service platform, which has been applied on a large scale in the fields of intelligent driving, mass consumption, public services, and smart cities, and has produced significant economic and social benefits.

Project Title: Key technologies and applications of efficient video enrichment based on spatiotemporal visual perception and multi-source information fusion

Completed units: Shijiazhuang Tiedao University, the 54th Research Institute of China Electronics Technology Group Corporation, Xiamen Pujing Intelligent Technology Co., Ltd

Main finishers: Zhang Yunzuo, Dong Yanlei, Wang Zhengyou, Huo Lei, Gao Jinpiao

Focusing on the technical problem of video condensation in open scenes, this project proposes the concept of spatiotemporal trajectory motion characterization model and object-oriented compact keyframe of visual perception energy, realizes adaptive frame-level video enrichment of motion state, and forms a set of efficient video enrichment theory and technical system based on spatiotemporal visual perception and multi-source information fusion, which has been applied in related enterprises.

Project Title: Large Model Lightweight Technology and Its Application in Life Service Internet Platform

Completed by: Beijing Sankuai Online Technology Co., Ltd. (Meituan)

Main finishers: Wang Jingang, Tang Hongyin, Yang Yang, Liu Jiahao, Li Ruyin, Cai Xunliang

Focusing on the difficulties and challenges of large-scale pre-training models in the Internet life service platform, this project designs a lightweight technical system for large models with high compression ratio requirements, proposes a large model adaptation method for text representation and retrieval, and builds a pre-training application platform to promote the results, which are widely used in life service business scenarios such as Meituan search, advertising, customer service and R&D efficiency, and achieve good social and economic benefits.

Project Name: Key technologies and applications of cloud-native service platform in multi-cloud management mode for government industries

Completed unit: Inspur Software Technology Co., Ltd

Main finishers: Li Cunbing, Chen Huanxin, Wu Di, Liu Jinge, Yang Jian, Lv He

Focusing on cloud native services, this project proposes unified adaptation, scheduling and hierarchical operation and maintenance operation and maintenance management technology based on multi-cloud/heterogeneous cloud resources, resource collaboration and cloud support technology for multi-form industry applications, and independently develops the key technologies of cloud native service platform in the multi-cloud management mode for the government industry, which are applied to Inspur industry digital platform, Inspur industry cloud service management platform and other products, and achieve significant economic and social benefits.

CCF recommended

【Featured Articles】

• 2023 "CCF Overseas Science and Technology Figure Award" selection results announcement
• 2023 "CCF Outstanding Education Award" selection results announcement