（来源：千龙网） 记者2月13日从上纬启元获悉，由北京大学教授董豪领衔的上纬启元团队取得重要突破，为机器人装 上"空间大脑"。 尽管今天的机器人能识别室内物品，却对"右边""后面""绕一圈"这样的方位词一头雾水。给机器人装 上"空间大脑"，一项名为NavSpace的研究成果已由北大、启元团队联合完成并发表于机器人领域国际顶 级会议ICRA 2026。 研发人员构建了NavSpace评测基准——业内首个专注于"空间智能"的导航测试集。在此基础上，研发出 SNav模型，能从现有数据中自动生成大量涵盖六大空间能力的训练样本，从而将空间感知与推理能力 高效注入模型。搭载该模型的机器人在真实环境中能听懂并执行如"去二楼最近的沙发旁"等复杂指令。 这项突破将直接推动智能服务机器人走进真实场景。 ...

"未来，机器人本体制造或将逐步成为行业准入门槛，真正决定具身智能产品能力的将是大脑。"原力无 限联合创始人刘扬在接受中国证券报记者专访时表示，原力无限发力机器人大模型，探索机器人处理复 杂长程任务、持续学习、理解事件因果的能力并建设AI基础平台，以此提升更多类型的具身智能产品 在应用场景中的表现。 2月4日，原力无限发布了其首款消费级具身智能机器人小原子，提出打造能不断进化的生命体。小原子 在硬件层面打造了通用外设接口，使其能替换手部和其他关键部件，实现功能扩展，并在软件层面通过 官方OTA和开放平台持续升级。 ● 本报记者 王婧涵 方悦 攻坚具身大脑 "机器人未来的竞争本质是大脑的竞争。"刘扬认为，中国拥有强大的制造业基础，随着行业成熟，未来 机器人本体可能逐渐演变为标准化程度较高的硬件载体，而真正的技术能力差异将集中于软件能力也就 是所谓的机器人大脑层面。 他表示，原力无限自创立之初，便将研发重心押注于算法模型与AI基础建设平台（AI Infra）两大方 向。在算法模型层面，原力无限通过聚焦三大维度提升机器人在不同场景中的表现。 首先，聚焦处理长程任务的能力，即机器人能将笼统复杂的用户指令自动分解为一系列 ...

当日，在黑龙江齐齐哈尔开往北京的T48次列车上，哈尔滨铁路部门连续二十年举办的列车"春晚"如约 而至。在多样的节目中，两个仿生机器人成为全场焦点，旅客沉浸在列车"春晚"温馨氛围的同时，也感 受到了年味中满满的科技感。 新华社图文 2月13日，小朋友在T48次列车上与仿生机器人握手。 ...

经济观察网基于最新公开信息，以下是特斯拉在2026年可能影响其业务和股价的值得关注事件。这些事 件均来自近期媒体报道和公司公告，但请注意时效性，具体进展请以官方发布为准。 战略推进：特斯拉与马斯克旗下的xAI和SpaceX正推进深度整合计划，包括投资协议和潜在合并讨论， 旨在构建覆盖太空AI、人形机器人和自动驾驶的生态体系。马斯克曾表示长期市值目标达100万亿美 元，但整合面临监管审查和技术挑战。 行业政策与环境：美国众议院委员会于2026年2月10日通过《SELF DRIVE Act》，为无方向盘的纯自动 驾驶车辆(如特斯拉Robotaxi和Cybercab)在全国范围内测试与商业化铺平道路，预计2027年前完成安全 标准制定。 高管变动：2026年2月，特斯拉北美销售及服务主管离职，延续了2025年以来的高管离职潮，涉及软 件、电池和机器人等关键领域，可能对运营稳定性带来影响。 业绩经营情况：根据2025年财报，特斯拉全球电动车交付量达163.6万辆，亚太市场创历史新高；储能 业务装机量同比增长48.7%。这些数据反映了基本面的韧性。 业务与技术发展：特斯拉自主研发的AI5和AI6自动驾驶芯片取得进展，计 ...

Group 1 - The core idea of the articles revolves around the advancement and application of AI robots in various sectors, particularly in enhancing visitor experiences in museums and improving operational efficiency in factories [1][2][3] Group 2 - The AI robot "Xiaoxin" has completed a six-month internship and is set to enhance visitor experiences during the Spring Festival by providing interactive and informative tours [1] - The robot is equipped with advanced features such as a high-fidelity design, laser radar, and visual navigation, allowing it to navigate crowds and interact effectively with visitors [1] - The robot's capabilities are supported by a data analysis module that records visitor interactions, providing valuable insights for optimizing exhibit layouts and content [1] Group 3 - The Chengdu robot training center, a collaboration between Tianfu Jiangxi Laboratory and Xinhua Innovation Center, focuses on developing robots for various real-world applications [2] - The center employs a "data + scenario + training" model, utilizing a 1:1 simulation system to replicate diverse environments for effective robot training [2] - The Chengdu government aims to deepen the "Artificial Intelligence +" initiative, promoting AI across multiple sectors to foster a smart economy and society [3]

Group 1 - The first Global Humanoid Robot Free Fighting League (URKL) was launched in Shenzhen, attracting global attention from the technology and sports sectors [3][5] - The championship team will receive a 10-kilogram gold belt valued at 10 million yuan, symbolizing the highest honor and recognition of technological prowess and innovation [5][6] - The event aims to push technology from the lab to practical applications, serving as a "touchstone" for innovation and talent discovery in the humanoid robotics field [6][7] Group 2 - The competition rules emphasize non-violent, technical contests, focusing on robot control, balance, perception, and decision-making under pressure [5][6] - The humanoid robotics market is entering a phase of large-scale production, with companies like Yushu Technology and Zhiyuan Robotics leading the way, each expected to exceed 5,000 units in actual shipments by 2025 [6][7] - The URKL serves as a platform for international technical exchange and competition, exploring the integration of technology, sports, and culture [7][8] Group 3 - The competition is designed to test the reliability and intelligence of humanoid robots in high-intensity environments, addressing challenges such as maintaining balance and executing tactical tasks under constraints [6][8] - The event is seen as a collective "stress test" for the humanoid robotics industry, transitioning from demonstration to practical applications [8]

Core Insights - Elon Musk discussed the economic benefits of space data centers, the challenges of scaling electricity generation on Earth, and the conditions needed for large-scale production of humanoid robots in the U.S. [1][2] Group 1: Space Data Centers and Energy Challenges - Musk emphasized that moving computing power to space is not primarily about saving on electricity costs but addressing the energy supply issue, as ground power generation cannot keep pace with the exponential growth of chip computing power [1][4] - He noted that solar energy efficiency in space is approximately five times greater than on Earth, and that deploying AI in space could become the most cost-effective solution within the next 30 to 36 months [6][7] - The bottleneck for expanding energy generation on Earth is not just the power plants but also the manufacturing capabilities for turbines and transformers, which are currently limited [10][11] Group 2: AI Deployment and SpaceX's Role - Musk predicts that in five years, the amount of AI deployed and operational in space will exceed the cumulative total on Earth, with annual AI capacity in space potentially reaching hundreds of gigawatts [23][26] - SpaceX aims to become a major supplier of computational power in space, potentially surpassing all terrestrial institutions combined [26][27] Group 3: SpaceX IPO and Funding Strategies - Musk indicated that discussions around a potential SpaceX IPO are increasing due to the need for substantial funding that exceeds what private markets can provide, suggesting that the public market offers significantly more capital [28][32] - The focus remains on speed and overcoming bottlenecks, with Musk stating that if funding is the only bottleneck, it will be addressed through an IPO [33][34] Group 4: Manufacturing and Supply Chain Challenges - Musk highlighted the critical need for manufacturing capabilities, particularly for turbine blades and solar panels, which are essential for energy generation in space [12][18] - He mentioned that SpaceX and Tesla are working towards achieving a solar capacity of 100 gigawatts, indicating a comprehensive approach to the solar supply chain [22][42] Group 5: AI and Human Interaction - Musk expressed concerns about the future relationship between AI and humanity, suggesting that as AI intelligence grows, human control may diminish, emphasizing the importance of instilling the right values in AI systems [49][50] - The ultimate goal is to ensure that AI contributes positively to the continuation of human civilization and understanding of the universe [51][52]

Core Insights - The article highlights the transformation of the construction machinery industry through the implementation of intelligent manufacturing and AI technologies, with Zhonglian Heavy Industry's smart factory leading the way [1][6]. Group 1: Intelligent Manufacturing and AI Integration - Zhonglian Heavy Industry's smart factory features over 280 industrial robots and nearly 300 AGVs, achieving an average output of one excavator every six minutes [1]. - The factory addresses the challenge of "multi-variety, small-batch" production in the construction machinery sector by utilizing a shared manufacturing model [2][3]. - AI technology is deeply integrated, with an application rate exceeding 80%, enabling self-organization and optimization of production resources [3]. Group 2: Production Efficiency and Cost Reduction - The smart factory has achieved high-efficiency mixed-flow production of over 100 excavator models, along with cross-product collaboration among excavators, cranes, and other machinery [3][4]. - The use of automated logistics has reduced in-process inventory by 70% and lowered construction costs by 15%, while steel plate material utilization exceeds 90% [4]. - The shift to a "sales-driven production" model allows for customized production based on client needs, significantly reducing finished goods and work-in-progress inventory [7]. Group 3: Market Impact and Future Prospects - Following its designation as a national pilot-level smart factory, Zhonglian Heavy Industry has seen a surge in orders, totaling 1.1 billion yuan from global clients [8]. - The smart manufacturing solutions developed at this factory have been replicated in over 20 other smart factories worldwide, enhancing collaboration across the entire industrial chain [8]. - The company anticipates that its humanoid robots will begin mass production and market sales by 2026, further expanding its technological capabilities [5].

Group 1 - Major automotive companies are accelerating their entry into the robotics sector, with Tesla, Chery, Hyundai, Xpeng, and Seres all making significant moves by 2026, including Tesla repurposing production lines for robots [2][4] - Chery's humanoid robot "Wuyou" has been integrated into local law enforcement in Wuhu, showcasing its capabilities in traffic management and marking a significant step in the application of robotics in public services [3][4] - The automotive industry is redefining its value through the development of embodied intelligence, moving beyond traditional automotive definitions and integrating "car, human, and machine" [2] Group 2 - Chery's humanoid robot has entered over 30 countries and is involved in various applications, indicating a strong growth trajectory for the company in the robotics field [4] - Seres is also expanding into embodied intelligence, establishing a new company focused on smart robotics and artificial intelligence, with a registered capital of 50 million RMB [5] - Tesla plans to produce 1 million units of its third-generation humanoid robot by the end of 2026, indicating a significant commitment to this new product line [7] Group 3 - Hyundai plans to deploy over 30,000 humanoid robots in its U.S. factories by 2026 to enhance production efficiency and reduce costs [8] - The trend of automotive companies venturing into robotics is spreading to upstream suppliers, with companies like Top Group and Ningbo Huaxiang actively exploring embodied intelligence [9] - CATL has launched a humanoid robot named "Xiao Mo" in its battery production line, marking a significant step in integrating robotics into complex industrial processes [10]

Core Insights - The robot leasing market is expected to experience explosive growth, with Qingtian Rental CEO indicating a significant increase in orders and the necessity of a city partner program [1][3][11] Group 1: Market Growth and Projections - Qingtian Rental has received over 1,000 orders covering the Spring Festival holiday, with expectations to exceed 5,000 orders post-holiday [1] - The market size for robot leasing is projected to grow from 1 billion yuan in 2025 to 10 billion yuan in 2026 [1][11] - Orders for robot leasing have shown a continuous month-on-month increase, with a 30% increase in the two weeks leading up to the Spring Festival and a projected 60% increase during the holiday week [3] Group 2: Challenges in the Industry - The robot leasing industry faces issues such as price transparency, lack of service standards, and uncontrollable satisfaction levels [5] - There is a mismatch in supply and demand, with some areas having demand but insufficient supply, while others have supply but lack efficiency [4] Group 3: Strategic Initiatives - Qingtian Rental has launched a national city partner strategy to recruit local partners, aiming to lower entry barriers through subsidies, technical support, operational training, and traffic protection [6][8] - The company emphasizes the importance of building a nationwide service network for the scalable application of robots [9] Group 4: Investment and Partnerships - Qingtian Rental is in discussions with investment teams from major companies like Alibaba, Tencent, and ByteDance, indicating strong interest from capital markets [10][12] - The company plans to pursue a financing rhythm of "one round per month" post-Spring Festival, aiming to complete a phase of financing by mid-2026 [12]