【环球网科技综合报道】近日，北京大学与高德达成战略合作，双方将围绕空间智能领域展开深度合作，重点开展技术联合攻关、智慧社区建设及教研实践 基地共建等工作，以"学术赋能产业，产业反哺科研"为核心理念，推动产学研深度融合，树立行业新标杆。签约仪式由北京大学副校长朴世龙、科技开发部 部长姚卫浩与高德集团董事长刘振飞、高德副总裁董振宁等共同出席。 北京大学作为我国顶尖的高等学府，在计算机科学、空间信息技术、人工智能等领域拥有深厚的科研基础和卓越的创新能力，尤其是在计算机视觉、大数据 处理等技术领域处于国际领先地位。这些优势与高德在空间智能领域的技术需求高度契合，为双方的合作奠定了坚实的基础。 在今年的云栖大会上，我曾提出一个观点："空间智能将像云计算一样，成为360行与物理世界交互的标配。" 这句话的背后，是高德对AI演进路径的判断： 语言模型让AI学会了"说话"，而空间智能，则要让它真正"看见"、"理解"，并"行动"。这样一来，AI在真实世界中才有了更深远的存在意义。 北京大学副校长朴世龙表示，2026年是"十五五"规划开局之年，此次签约，既是双方主动对接国家战略、抢抓"十五五"发展机遇的务实之举，更是携手在空 间 ...

Core Viewpoint - Gaode has announced a strategic partnership with Peking University to collaborate in the field of spatial intelligence, focusing on technology joint efforts, smart community construction, and the establishment of a research and teaching practice base [1] Group 1: Academic Research - Gaode is collaborating with Peking University's Wang Xuan Computer Research Institute to focus on cutting-edge directions in spatial intelligence, aiming for theoretical breakthroughs in core technologies such as multi-source data fusion and dynamic 3D reconstruction through joint projects and resource sharing [1] Group 2: Business Practice - Based on existing project outcomes, both parties will explore application scenarios and commercialization paths in emerging fields such as 3DGS & 4DGS technology, embodied intelligence, and low-altitude economy [1] - The partnership plans to jointly promote the construction of Peking University's "Friendly Community," applying spatial intelligence technology to practical scenarios such as campus transportation and smart services [1] Group 3: Talent Development - Both parties will establish a spatial intelligence research and teaching base to create a platform for talent cultivation and technical exchange [1]

Group 1 - The core plan for 2026 includes the implementation of the new Jingxi Action Plan and the "Two Parks and One River" initiative, focusing on projects like the Artistic Waterfront and the Lianshi Lake Water Sports and Leisure Base [1][3] - The "Two Parks and One River" initiative encompasses Shijingshan's Shougang Park, Fengtai Garden Expo Park, and the Yongding River, aiming to create a new development hub in the western part of Beijing [3] - Shijingshan District will enhance urban design, focusing on waterfront spaces and greenways, and will promote key projects such as the West Mountain Yongding River Cultural Belt Greenway and Shougang's surrounding landscape lighting [3] Group 2 - The district plans to expand diverse consumption formats in culture, tourism, and commerce along the Yongding River, including projects like the second phase of the Trendy Enjoyment Waterfront and the Lianshi Lake Water Sports Leisure Base [3][4] - The district aims to establish a permanent venue for the Service Trade Fair, enhancing venue functionality and surrounding infrastructure to create an international exhibition town [4][5] - Key projects like the Shougang North District International Talent Community and the overall enhancement of the Shougang Ski Jump area will be prioritized to improve living and working conditions [5]

作者｜黄楠 编辑｜袁斯来 硬氪获悉，深圳市无穹创新科技有限公司（以下简称"无穹创新"）近期连续完成三轮融资，累计金额达数千万元，Maple Pledge枫承资本长期出任独家私募 融资顾问。我们总结了本轮融资信息和该公司几大亮点： 融资轮次：天使+轮、Pre-A及Pre-A+轮 融资规模：数千万元 投资方：李泽湘教授旗下的XBOTPARK基金、零以创投、元禾原点广州基金、元禾原点杭州基金、金沙江联合、晨晖创投及甘洁教授旗下的知行基金 资金用途：将主要用于技术研发投入、核心团队人才扩充与全球商业化拓展 融资金额及投资机构 公司基本信息 成立时间：2020年12月 注册地址：深圳市南山区 企业定位：无穹创新是一家聚焦空间智能基础设施的硬科技企业，聚焦全自研跨本体空间智能通用硬件MetaCam，为人、机器人及智能系统提供与空间深度 交互的感知与理解能力。 技术亮点：基于"通用硬件+整机产品"的双业务布局，无穹创新构建了从算法层、硬件层到整机层的完整空间智能体系。 算法层面，公司自研并开源了一套多传感器融合算法，覆盖空间定位、环境感知、语义理解、决策规划与自主控制等环节。通过海量真实场景数据的持续训 练与优化，其跨场景 ...

Core Insights - The article discusses the evolution and significance of spatial intelligence in artificial intelligence (AI), emphasizing its potential to reshape creativity, embodied intelligence, and societal progress. It highlights the limitations of current AI technologies, particularly in understanding and interacting with the physical world, and proposes the development of "world models" to bridge this gap [1][2][3]. Group 1: Spatial Intelligence and AI Development - Spatial intelligence is a fundamental aspect of human cognition that influences how individuals interact with the physical world, from everyday tasks to complex problem-solving [7][8]. - Current AI technologies, particularly large language models (LLMs), have made significant advancements but still lack the spatial reasoning and understanding necessary for real-world applications [12][13]. - The development of AI with spatial intelligence is seen as the next frontier, requiring new generative models that can understand, reason, and interact with both virtual and real environments [14][19]. Group 2: World Models and Their Capabilities - World models are defined as generative models that possess three key capabilities: generative, multimodal, and interactive [14][16][17]. - These models must be able to generate consistent simulations of the world, process various forms of input, and predict the next state based on actions taken [14][16][17]. - The creation of effective world models is essential for advancing AI's spatial intelligence and enabling applications across various fields, including robotics, creative industries, and scientific research [19][30]. Group 3: Applications and Future Directions - The potential applications of spatial intelligence in AI span multiple domains, including creativity, robotics, and healthcare, with the aim of enhancing human capabilities rather than replacing them [22][31][32]. - In creative fields, tools like the Marble platform are being developed to empower creators by allowing them to build immersive and interactive environments without the constraints of traditional design processes [24][25]. - In healthcare, spatial intelligence can transform processes from drug discovery to patient monitoring, enhancing the quality of care while maintaining human connections [32][33].

转自：半月谈网 近日，高德地图助力清华大学打造的国内首个AI智慧校园——"友好社区"项目正式落地。该项目以AI技术为核心，依托高德空间智能架构，推出了一系列智 能化出行服务，实现校园交通管理与师生出行体验的双重升级，为全国高校智慧校园建设探索出一条可复制、可持续的校企公益协作路径。 作为"友好社区"项目的重要成果之一，高德国内首个校园公交导航功能已正式上线，显著提升了师生与访客的出行体验。用户只需在高德地图中输入校内起 终点（如"清华大学西门→新清华学堂"），选择"公共交通"，即可一键查看校园巴士的线路走向、停靠站点，并实时获取车辆位置和预计到站时间。这一功 能让"卡点候车"成为现实，大幅减少了等待时间。同时，还能自动规划最优出行方案，实现校园巴士与校外公交、地铁的无缝衔接。 在智慧校园实践中，高德首次实现了校园内实时路况的精准上线。通过AI技术的深度应用，高德地图针对校园高峰期主要道路和出入口易拥堵的痛点，对 清华校园内多条道路、多类通行规则及限行区域等累计120余条信息进行了全面梳理和精准标注。师生与访客在使用导航时，能够实时查看拥堵路段并获得 推荐路线，从而有效避开高峰潮汐拥堵，显著提升出行效率。这一创 ...

Core Insights - Gaode Map has transformed from a traditional navigation tool into a comprehensive service platform, achieving 950 million monthly active users as of September [1][5] - The company has diversified its offerings to include ride-hailing, hotel bookings, and local services, positioning itself as a "super entrance" for various lifestyle needs [3][17] - Despite its growth, Gaode faces challenges in monetization and user experience, struggling with high operational costs and user retention [5][16] Group 1: User Growth and Engagement - Gaode Map has reached 950 million monthly active users, making it one of the top six apps in China [1] - The app's daily active users (DAU) surpassed 100 million by 2018, becoming the first travel platform in China to achieve this milestone [5] - User engagement remains a challenge, as the average usage time is significantly lower compared to leading competitors [5][6] Group 2: Business Model and Revenue Streams - Gaode's revenue heavily relies on online advertising, accounting for 65% of its income, with additional revenue from navigation services and ride-hailing commissions [6] - The company has shifted from a paid model to a free service model since 2013, which has impacted its profitability [4] - In the 2025 fiscal year, Gaode achieved profitability for the first time by reducing costs [6] Group 3: Service Diversification and Innovation - Gaode has expanded its services to include car rentals, hotel bookings, and local dining recommendations, aiming to create a comprehensive lifestyle platform [3][9] - The launch of "Gaode Guide" and "Gaode Street List" has positioned Gaode as a competitor to existing review platforms, utilizing real user behavior data for more accurate recommendations [11][12] - The company is exploring partnerships, such as with XPeng Motors for Robotaxi services, to enhance its service offerings [17] Group 4: Challenges and User Experience - Gaode faces operational challenges, including regulatory compliance issues and user complaints regarding service quality and customer support [12][13] - The introduction of targeted advertising has received mixed feedback, with some users expressing dissatisfaction over intrusive ads during navigation [16] - Balancing user experience with monetization strategies remains a critical challenge for Gaode as it seeks to expand its service portfolio [16][18]

Core Viewpoint - The article emphasizes the transition of autonomous driving from "perception-driven" to "spatial intelligence" by 2025, highlighting the importance of understanding and interacting with the three-dimensional physical world [3]. Group 1: Spatial Intelligence Definition - Spatial intelligence is defined as the ability to perceive, represent, reason, decide, and interact with spatial information, which is crucial for the interaction between intelligent agents and the physical world [3]. - Current spatial intelligence is primarily focused on perception and representation, with significant room for improvement in reasoning, decision-making, and interaction capabilities [3]. Group 2: World Models and Simulation - GAIA-2 is a multi-view generative world model for autonomous driving that generates driving videos based on physical laws and conditions, addressing edge cases in driving scenarios [5]. - GAIA-3 enhances GAIA-2 by increasing the scale fivefold and capturing fine-grained spatiotemporal contexts, representing the physical causal structure of the real world [9]. - ReSim combines expert trajectories from the real world with simulated dangerous behaviors to achieve high-fidelity simulations of extreme driving scenarios [11]. Group 3: Multimodal Reasoning - The SIG framework introduces a structured graph scheme that encodes scene layouts and object relationships, aiming to enhance geometric reasoning in autonomous driving [16]. - OmniDrive generates a large-scale 3D question-answer dataset to align visual language models with 3D spatial understanding and planning [19]. - SimLingo addresses the alignment of driving behavior with semantic instructions through an action dreaming task, demonstrating the potential of general models in real-time decision-making [21]. Group 4: Real-time Digital Twins - DrivingRecon is a 4D Gaussian reconstruction model that predicts parameters from surround-view videos, enabling efficient dynamic scene reconstruction for autonomous driving [26]. - VR-Drive enhances robustness in driving systems by allowing real-time prediction of new viewpoints without scene optimization [29]. Group 5: Embodied Fusion - MiMo-Embodied is the first open-source cross-embodied model that integrates autonomous driving with embodied intelligence, showcasing significant transfer effects in spatial reasoning capabilities [31]. - DriveGPT4-V2 is a closed-loop end-to-end autonomous driving framework that outputs low-level control signals, evolving from visual understanding to closed-loop control [36]. Group 6: Industry Trends - By 2025, the industry is moving towards an end-to-end VLA architecture, leveraging large language models for driving decision-making [40]. - Waymo's EMMA model integrates multimodal inputs and outputs in a unified language space, enhancing complex reasoning in driving tasks [41]. - DeepRoute.ai's DeepRoute IO 2.0 architecture introduces chain-of-thought reasoning to address the "black box" issue in end-to-end models, improving user trust in autonomous systems [44].

Core Insights - The article discusses the increasing influence of AI technology and the role of spatial intelligence in various industries, highlighting the efforts of Qunkong Technology in integrating 3D data with AI capabilities for applications in design, digital twins, robotics training, and content creation [1][3]. Product Matrix Development - Qunkong Technology has accelerated the release of new products, including the 3D AI content creation tool LuxReal, aimed at e-commerce professionals and filmmakers, which efficiently generates video content while maintaining spatial consistency [2]. - The company introduced several platforms, such as SpatialVerse for interactive training data, SpatialTwin for industrial AI applications, and an AI design platform on Cool Home, which has launched an overseas version targeting markets in South Korea, Thailand, and Europe [2]. Strategic Vision - CEO Chen Hang emphasized the importance of data in AI applications, noting the scarcity of real data in the spatial intelligence field and the company's development of a data "flywheel" over the past 14 years [3]. - Chairman Huang Xiaohuang stated that the company is focusing on building core capabilities and is open to external collaborations, indicating a commitment to long-term growth across various sectors [3]. Industrial Scene Exploration - At the World Internet Conference, Qunkong Technology unveiled the SpatialTwin platform, which aims to enhance digital twin systems in factories as more intelligent robots are integrated [4]. - The platform is designed to utilize high-density simulation computing power and facilitate real-time collaboration among participants, improving operational efficiency and reducing costs in industrial settings [5]. Expansion into Embodied Intelligence Market - Qunkong Technology is broadening its business scope from spatial design to content creation and embodied intelligence, positioning itself as a provider of spatial intelligence infrastructure [6]. - Collaborations with companies like Moxianfei and Digua Robot aim to innovate robot simulation training solutions, addressing the challenges posed by diverse hardware and data formats in the industry [6]. Future of Training Data - Huang Xiaohuang noted that some clients currently use a 9:1 ratio of simulation data to real data, with expectations that the reliance on digital training will increase as technology evolves [7]. - The company believes that spatial intelligence is crucial for the future development of the robotics industry, enabling robots to understand and operate effectively in the physical world [7].

Core Insights - The article highlights the increasing importance of spatial intelligence technology in various industries, driven by advancements in AI technology. The company, Qunkong Technology, is integrating its accumulated 3D data with AI to enhance capabilities in areas such as spatial design, industrial digital twins, robotic simulation training, and content creation [1][2]. Product Matrix Development - Qunkong Technology has accelerated the release of new products, including the 3D AI content creation tool LuxReal, aimed at professionals in e-commerce, film, and advertising. This tool efficiently generates video content while maintaining spatial consistency, addressing common issues in AI-generated videos [1][2]. - The company also introduced several platforms, including SpatialVerse for intelligent training, SpatialTwin for industrial AI twins, and an AI design platform on Cool Home, which has launched an overseas version targeting markets in South Korea, Thailand, and Europe [2]. Industrial Scene Exploration - The SpatialTwin platform focuses on the application of AI in factory automation, allowing for real-time collaboration and information sharing among participants. It utilizes high-density simulation computing power to enhance operational efficiency and reduce costs in industrial settings [3][4]. - Qunkong Technology has partnered with Hangcha Group and Stand Robot to develop integrated intelligent logistics solutions and promote the large-scale application of industrial embodied intelligence [4]. Expansion into Embodied Intelligence Market - The company is broadening its business scope from spatial design to content creation and embodied intelligence, positioning itself as a provider of spatial intelligence infrastructure. Collaborations with Motphys and Digua Robot aim to innovate robotic simulation training solutions [5]. - The industry faces challenges in training data for embodied intelligence robots, with companies employing both real-world training and simulation platforms to enhance training efficiency. Qunkong Technology's clients are increasingly using synthetic data, with some reporting a 9:1 ratio of synthetic to real data [5].