另一项成熟应用是城市交通信号优化。 王凯指出，由大模型驱动的信控智能体，已能自主生成并下发 红绿灯配时方案。该应用从广州珠江新城十几个路口的试点，已扩展至六个主城区，实现区域级24小时 不间断优化，有效提升了通行效率，并已推广至重庆、西安等城市。 当强对流天气即将来袭，AI已提前"读"懂天气与舆情，为机场调度好应急资源；当你驾车经过路口，红 绿灯配时可能正由云端"智能体"实时优化。12月12日，2025年粤港澳大湾区人工智能与机器人产业大会 （XAIR）——"AIR+交通"专场活动暨人工智能赋能新型工业化供需对接深度行（广州站）活动上，专 家指出，交通智能化已不止于"感知反应"，更在于"预测决策"。场景丰富的广东，正成为实践这一前沿 理念的核心阵地。 省委统战部副部长、省工商联党组书记陈志清在致辞中表示，"广东智造"的交通解决方案，正从湾区走 向全国、走向世界，佳都科技等优秀企业在智慧轨道、智慧交通等领域快速成长。他表示，期待广大民 营企业勇担重任、走在前列，发挥示范引领作用。省工商联将继续发挥桥梁纽带功能，积极搭建服务平 台，支持民营企业加强科技创新，推动智能交通产业高质量发展。 "智慧交通的未来必将迈向' ...

Core Insights - Jiaodu Technology (600728) has initiated its listing plan on the Hong Kong Stock Exchange, marking the beginning of its "A+H" listing journey, with joint sponsors including GF Securities, CITIC Securities, and Agricultural Bank of China International [1] - The company focuses on smart rail transit and smart city transportation, leveraging AI, big data, and digital twin technologies to enhance urban transportation [1] - The A+H listing aims to broaden overseas financing channels, enhance international brand influence, and support the internationalization of its AI technology and smart transportation business [1] Company Developments - Since its first project in 2004 with the Guangzhou Metro Line 4 automatic ticketing system, the company has achieved significant technological breakthroughs and deployed smart transportation solutions in over 100 cities, including Shanghai, Guangzhou, and Chengdu [1] - As of June 30, 2025, Jiaodu's intelligent products and services will cover 45 cities and over 2,200 metro stations, contributing to the digital and intelligent upgrade of more than 4,100 kilometers of metro networks nationwide [1] Technological Innovations - The company has developed the world's first large model for the transportation industry, "Zhixing Model," which supports complex applications such as real-time network status prediction and dynamic signal control optimization [1] - Jiaodu Technology has introduced the first domestic open-source operating system for the transportation industry, "Jiaohong," which facilitates seamless collaboration and computing power sharing among various transportation devices [2] Future Plans - According to the prospectus, the company plans to increase investment in R&D, focusing on AI-driven edge intelligence systems and city-level multimodal transportation digital twin platforms [2] - The company also aims to strategically invest in and acquire technologies in the smart transportation and advanced AI sectors to enhance its technological capabilities and global competitiveness [2]

Core Insights - The fourth Zhuhai Algorithm Competition showcased the latest advancements in AI applications, emphasizing the importance of practical implementation in various industries [5][16][18] - Over 8,000 teams participated in the competition, highlighting the growing interest and innovation in AI technologies [5][18] Industry Applications - The education sector is a prominent area for AI model applications, with companies like iFLYTEK introducing innovative models such as the "Spark Classroom Analysis Model," which provides comprehensive teaching analysis [7][8] - AI models are also being utilized in marketing, with companies like Kuaizi Technology developing platforms for commercial video production, achieving significant growth and reaching over 8,500 brands [10] - In the safety sector, AI models are being integrated for risk management, with companies like Guangzhou Dimension Breakthrough Technology focusing on providing new safety solutions [12][11] Data Utilization - Data has become a valuable resource in the digital economy, with companies like Guangzhou Linglingqi Technology leveraging proprietary data to develop specialized AI models for the automotive repair industry [13] - The "Pushi AI Nutritionist" model by Pushi Technology is the first approved model in the smart cafeteria sector, analyzing student dietary intake and ensuring nutritional balance [14] Algorithm Innovation - The Zhuhai Algorithm Competition has become a significant driver of algorithm innovation, focusing on real-world challenges faced by enterprises [16][17] - The competition has seen a rise in participation, with over 8,131 teams from more than 30 countries competing this year, reflecting its growing influence [18]