Core Viewpoint - Tesla has made significant progress in the field of autonomous driving with its "full self-driving" (FSD) system, showcasing a fully automated delivery of a Model Y without human intervention, which is considered a major milestone in the development of autonomous driving technology [1][2]. Technical Performance - The Model Y's delivery route included complex suburban roads and residential areas, achieving Level 4 (L4) autonomy, which requires no human intervention under specific conditions, and surpassing the performance of Waymo's previous tests [1][2]. - The FSD system utilizes a custom neural network chip and a comprehensive perception system that includes cameras, LiDAR, and ultrasonic sensors to navigate various driving scenarios without human assistance [2]. Industry Trends - The excitement around autonomous driving has cooled, with major companies like Ford and Volkswagen shutting down their autonomous driving ventures, and Apple terminating its car development plans [3][4]. - General Motors closed its Cruise Robotaxi project after an incident involving a pedestrian, redirecting resources towards consumer-oriented autonomous vehicle development [4]. Challenges to Full Autonomy - Achieving Level 5 (L5) autonomy remains challenging due to the need for systems to handle all possible driving conditions and scenarios, which requires extensive sensors and complex computing platforms [5]. - Public and regulatory tolerance for machine errors is low, despite autonomous vehicles having a lower overall accident rate compared to human drivers, complicating the path to widespread adoption [6]. Shift Towards Advanced Driver Assistance Systems (ADAS) - Advanced Driver Assistance Systems (ADAS) at Level 2 (L2) and Level 3 (L3) are becoming more viable, with features like automatic parking and lane-keeping gaining traction in the market [6][7]. - Companies like Mercedes and Volvo are integrating advanced safety features into their vehicles, while Chinese automakers like BYD are promoting the adoption of intelligent driving systems [7].

Core Viewpoint - The 2025 Global Digital Economy Conference opened in Beijing, where the United Nations Industrial Development Organization and Dongbi Technology Data Co., Ltd. jointly launched the global scientific literature data platform dbdata.com, which aims to enhance high-quality scientific research through AI integration and a proprietary database of nearly 90 million high-quality scientific literature metadata [1][2]. Group 1 - The dbdata.com platform is independently created by Dongbi Technology Data, featuring complete intellectual property rights and aims to break the Western monopoly in the field of scientific literature data [1]. - The platform is based on the "Dongbi Index," a new evaluation system from a Chinese perspective, which was released earlier this year and represents the first global academic journal evaluation system from China [1][2]. - The Dongbi Index employs a novel methodology that includes "seed journal citation tracing" and a hierarchical structure of citation networks, allowing for a more objective and accurate reflection of a journal's authority and influence in academia [2]. Group 2 - The platform categorizes journals according to the habits of the Chinese scientific and educational communities, enhancing its relevance and usability [2]. - Two AI features have been introduced on the platform: "Scholar Achievement Interpretation" and "Automatic Review Generation," with basic functionalities available for free to all users [2].

Core Insights - The article discusses the development of a solar energy project in Gansu, China, which utilizes a 260-meter high tower surrounded by over 12,000 heliostats to generate continuous electricity through a molten salt tower solar thermal power plant [1] - The project aims to leverage abundant solar energy for large-scale hydrogen production, with a focus on achieving industrial-scale hydrogen production through advanced electrolysis technology [1][2] - The establishment of the Hydrogen Energy and Low Carbon Center at Lanzhou University is a strategic move to explore various energy utilization pathways and build a hydrogen energy development community in the western region of China [1][2] Group 1 - The solar thermal power plant can generate electricity 24/7, enabling efficient and large-scale hydrogen production through water electrolysis [1] - Lanzhou University has developed an alkaline water electrolysis system capable of producing 1,000 cubic meters of hydrogen per hour, with an electrolysis efficiency exceeding 80% [1] - The "liquid sunlight" technology, which converts hydrogen and carbon dioxide into methanol, is highlighted as a viable solution for industrial carbon reduction and efficient hydrogen storage and transportation [2] Group 2 - The successful implementation of the world's first large-scale methanol synthesis demonstration project in Gansu and the upcoming 100,000-ton "liquid sunlight" fuel synthesis project in Inner Mongolia mark significant advancements in this technology [2] - Lanzhou University is positioned to leverage its interdisciplinary strengths in chemistry, physics, and energy to innovate in photovoltaic power generation, hydrogen production materials, and new energy storage technologies [2][3] - Future efforts will focus on flexible power generation, safe hydrogen-electric integration, and low-carbon energy utilization, with plans to establish a comprehensive research and testing platform for solar energy technologies [3]

Core Insights - The Shanghai Hongxin Drainage Pump Station's initial rainwater regulation and storage pipeline project has achieved a significant milestone with the launch of the country's first large-diameter initial rainwater regulation vertical shaft tunneling machine [1][2] - This project is a landmark in domestic mechanical vertical shaft construction, designed to intercept 4,700 cubic meters of initial rainwater per event, thereby alleviating pressure on urban drainage systems and improving downstream water quality [2] Group 1 - The project is located in Qibao Town, Minhang District, Shanghai, featuring a circular vertical shaft with a maximum excavation diameter of 16.3 meters and a depth of 44.6 meters [1] - The tunneling machine weighs approximately 522 tons and has a maximum power of 1,400 kilowatts, equipped with advanced intelligent tunneling technologies to address various construction challenges [1][2] Group 2 - The construction team employs a "data feedback + mechanical adjustment" dual control mechanism to ensure the verticality deviation of the deep shaft is less than 0.2% [2] - To tackle the challenge of the sinking of the shaft, the team enhanced the pushing cylinder's power to assist in the sinking process, thereby improving tunneling efficiency [2]

Core Insights - SportsGPT is the first large model for sports training and rehabilitation in China, providing analysis and personalized training plans based on user-uploaded videos [1][4] - The application has been recognized as a typical case in intelligent sports by the Ministry of Industry and Information Technology and the General Administration of Sport of China for 2024 [1] Sports Training Applications - SportsGPT offers real-time action guidance and personalized training plans, acting as a versatile assistant for sports training [4] - Users can upload videos to receive detailed breakdowns of their movements, identify errors, and obtain corrective suggestions along with a customized training plan [2][4] Rehabilitation Support - SportsGPT assists users with injury recovery by providing professional advice on treatment options and rehabilitation exercises [5][6] - The model can help rehabilitation specialists by collecting preliminary information and analyzing data to enhance treatment efficiency [5][6] Sports Research Efficiency - In traditional sports research, data collection and analysis are time-consuming; SportsGPT significantly improves efficiency by automating these processes [7] - The model learns from a vast dataset of elite athletes' biomechanics to quickly assess new athletes' performance and potential [7] Future Development - The next step for SportsGPT is to evolve from an efficient analyst to a predictive tool, capable of forecasting injury risks and performance fluctuations [8][9] - Achieving this requires integrating more physiological and psychological data to develop advanced algorithms [9]

在青海省海东市乐都区雨润镇羊圈村的冷凉蔬菜基地里，309个温棚整齐排列。种植户穿梭于温 棚，忙着给青椒浇水、吊线。 "温控器电压、温度值都正常，用电设备都检查完了，如果用电方面遇到什么问题，您随时给我们 打电话，我们马上过来检查。"不久前，国网乐都区供电公司"石榴籽"供电服务网格员盛廷成和同事张 学杰再次来到冷凉蔬菜基地检查用电设备，为丰收做好准备。 羊圈村冷凉蔬菜基地因地势较高，缺乏稳定水源，温棚灌溉一度成为难题。2024年7月，羊圈村党 支部书记唐生和向国网乐都区供电公司申请用电。国网乐都区供电公司总经理张永科了解情况后，立即 组织专班人员现场勘察并制定供电方案，组立杆塔108基，架设0.4千伏线路17.023千米，安装100千伏安 变压器4台，让蔬菜基地成功用上了稳定的泵水灌溉设备。今年5月，随着基地扩产，原来的灌溉设备无 法满足现在的用水需求，要再增加一台45千瓦的抽水设备。唐生和再次向国网乐都区供电公司申请增 容，并得到了帮助。 科技日报记者 张蕴 通讯员 赵梅芳 "现在合闸就能抽水灌溉，温控设备实现全自动化，产量不断增加。"唐生和介绍，基地采用电气化 设备后，整体产能提升60%，带动周边50余户村 ...

Group 1 - "Mosu Space" is rapidly emerging as a key AI innovation ecosystem in Shanghai, attracting over 100 companies and facilitating the establishment of nearly 400 AI firms in the Xuhui District, accounting for approximately 61% of the city's registered large models [1][2] - The ecosystem has developed a complete industrial chain from infrastructure construction to vertical applications in sectors such as consumption, finance, and entertainment [1] - The area aims to become the world's largest AI incubator, addressing challenges faced by AI companies through five public service platforms, including computing power scheduling and financial services [2] Group 2 - Companies like Wunwen Xinqun have significantly expanded their teams from 7 to nearly 200 members, with over 70% being R&D personnel, focusing on creating standardized computing power service platforms [2] - The AI industry is seeing practical applications with companies like Zhibo Zhixing and Xingji Meizu developing products that enhance human-computer interaction, such as intelligent cockpit systems and smart wearable devices [3] - Xuhui District plans to continue expanding the AI industry layout and support more AI companies with subsidies and assistance, while exploring independent operations and joint incubation models [3]

Group 1 - The Sichuan Aba Hongyuan Yake Photovoltaic Power Station is the first grid-connected energy storage power station in Sichuan Province, covering an area of 8,000 acres, equivalent to 784 standard football fields [1] - The project has a total installed capacity of 250 megawatts, with an expected annual power generation of approximately 450 million kilowatt-hours, saving 150,000 tons of standard coal and reducing carbon dioxide emissions by about 400,000 tons annually [1] - The project team overcame extreme environmental challenges, completing a 20-kilometer construction road in just 15 days and utilizing innovative construction techniques to enhance efficiency [2] Group 2 - The construction involved over 1,500 workers and more than 240 machines during peak periods, with the use of photovoltaic robots increasing installation efficiency by four times [2] - The design of the power station accommodates local pastoral needs, allowing for a "pastoral-photovoltaic complementary" model, with a minimum height of 2 meters for the solar panel supports [2] - The project includes the construction of intelligent warm sheds for highland meadows, achieving a survival rate of 95% for the stripped meadow area, minimizing disruption to local grazing lands [2]

Core Viewpoint - The integration of VR and LBE technology in cultural tourism is creating new immersive experiences, exemplified by the "No-Face God - Dazu Rock Carvings Take You Back to the Southern Song" VR experience in Tianjin, which has become a new hotspot for cultural consumption [1] Group 1: Technology and Innovation - The VR experience allows visitors to immerse themselves in the historical context of the Southern Song dynasty, with 1:1 precision digital replicas of classic sculptures like the Thousand-Handed Guanyin and Six Paths of Reincarnation [1] - The project utilizes LBE technology to support up to 20 participants simultaneously, enhancing the collaborative exploration of history [1] - The company has integrated core technologies such as software development, system integration, and cloud computing to create a seamless multi-user experience [1] Group 2: Cultural Significance - The project emphasizes the deep cultural connotations of the Dazu Rock Carvings, which are recognized as one of the world's eight major grottoes, reflecting ancient aspirations for a better life [1] - Historical accuracy is prioritized in the design, with attention to the tools used by artisans and the carving process, ensuring that every detail withstands scrutiny [1] - Engaging storylines are incorporated into the experience to enhance visitor engagement and interest [1] Group 3: Economic and Developmental Impact - The successful implementation of immersive cultural consumption scenarios offers new ideas for cultural heritage protection and the development of new consumption models [1] - The fusion of technology and culture is continuously generating innovative consumption scenarios, injecting vitality into urban development in Tianjin [1]

除了真实度，医学数字人核心医疗应用场景的另一大优点就是病例资源丰富，并且这些资源还会持 续更新和扩充。左铭说："在医学数字人核心医疗应用场景下，实习医生可以在正式进行临床实习前积 累大量的实践经验。" 视—听—触交互反馈式元宇宙智慧虚拟手术系统上不仅有10余种手术训练模式，还配备了个性化课 程定制及评分系统，帮助医生快速识别和改进操作中的问题。 通过突破数字虚拟人构建中的计算医学、融合建模、反馈交互等关键问题，上海交通大学医学院附 属瑞金医院正致力于推动医学数字人的全面发展，打造更加丰富的医学数字人核心医疗应用场景。 "一般的数字人，例如直播带货的虚拟主播，仅有人的外观，但医学数字人还要拥有虚拟的生理系 统，从而能够模拟患者的身体内部状况。"左铭说。这是上海交通大学医学院附属瑞金医院打造医学数 字人核心医疗应用场景的另一个目标：未来，不仅外科手术能在数字人身上先行先试，创新的内科治疗 方法也有机会在数字人身上先行验证。 坐在手术机器人控制台前，记者就能看到虚拟病人腹腔内部的脏器。借助手柄，操作者可精准执行 拉扯、剪切等手术动作，同时系统画面上还会显示虚线、提示语等，引导操作者操作…… 这套视—听—触交互反馈式 ...