近年来，无人机在农业、灾害管理及监视等领域展现出广泛应用潜力，其灵活性与数据采集能力为复杂任务提 供了有效支持。然而，着陆环节仍是操作中的关键难点， 据统计，超过47%的无人机事故发生于着陆阶段。 在强风、不平坦地形或移动平台等复杂环境下，着陆难度进一步增加。 当前多数小型多旋翼无人机采用刚性起落架，需在水平面上以较低速度着陆，以避免弹跳或翻倒。在高速移动 平台（如地面车辆）上着陆时，无人机常需保持较大俯仰角以抵抗空气阻力，而传统起落架设计难以适应此类 姿态。相关研究表明， 部分机型在特定条件下可实现最高约 50公里/小时的移动着陆，但受限于结构和控制 逻辑，成功率有限。 为提升着陆可靠性，部分研究引入磁性吸附机构或阻尼结构以增强着陆稳定性，并尝试通过提高下降速度以降 低风扰与跟踪误差的影响。然而， 在高速移动中着陆仍面临螺旋桨干涉、滑移及视觉跟踪失效等问题。 ▍融合摩擦减震与反向推力，突破无人机高速着陆技术瓶颈 该起落架采用摩擦减震器以消散着陆时的动能，有效防止无人机在冲击过程中发生反弹；同时通过反推力机制 增强脚部与着陆面之间的摩擦力，抑制滑移现象。这一策略使无人机能够以较高速度下降，并在接触前执行简 化 ...

Group 1 - On October 9, Yundongchu Technology launched the world's first industry-grade all-weather humanoid robot DR02, which meets IP66 waterproof and dustproof standards and can operate in temperatures ranging from -20℃ to 55℃, making it suitable for outdoor environments with rain and dust [2] - The DR02 has a height of 175cm and an arm length of 68cm, with good mobility to navigate complex terrains, capable of handling a load of 10kg with its arms and a total weight of 20kg [2] - The robot features a modular quick-release design, equipped with high computing power units and a multi-sensor system [2] Group 2 - Figure AI released its third-generation humanoid robot Figure 03 on October 9, which has a more lifelike appearance with a gray fabric covering, a 9% reduction in weight, and a smaller volume [5] - The Figure 03 includes a toe and arch design in its feet and utilizes wireless charging at the sole, showcasing five outfit options and a camera at the palm root for better object grasping assessment [5] - Despite competitors questioning its originality, Figure AI emphasizes significant differences in mechanical structure and aims to develop versatile products for both consumer and commercial markets [5] Group 3 - The "FutureBOT Future Gravity" 2025 Beijing Robot Culture Festival recently concluded, featuring immersive experiences across four future scenarios to promote the transition of robotic technology from laboratories to industries [8] - The event gathered leading companies such as Songyan Power and Stone Technology, showcasing technological innovations through physical displays and competitive performances, accelerating the implementation of innovative projects [8] Group 4 - The Qiaofei Cloud Health Care Robot was officially unveiled at the first AI Health Care Robot Achievement Exchange Conference held in Hunan on October 9, developed by a team of academicians and PhDs [11] - This robot integrates advanced sensing technology and AI algorithms, capable of accurately identifying body pain points within 5 seconds and generating a quantitative report in 3 minutes, providing a fully automated service [11] - It combines traditional Chinese medicine theories with AI technology to address standardization issues in health service, filling a gap in the health service market [11] Group 5 - Researchers at Carnegie Mellon University developed a new engineering method to create cilia-driven micro-bio robots called AggreBots using human lung cells [14] - This innovative approach utilizes cilia for movement instead of muscle fibers and introduces a modular assembly strategy, allowing for precise control of the robot's movement patterns [14] - AggreBots are biodegradable and biocompatible, with potential future applications in medicine, utilizing patients' own cells to avoid immune rejection [14]

Core Insights - IROS 2025 will be held in Hangzhou, China, from October 19 to 25, 2025, marking the second time this prestigious conference is hosted in China since 2006 [3] - The event aims to showcase cutting-edge advancements in intelligent robotics and automation, gathering over 6000 experts, scholars, and enterprises from around the globe [3] - A new initiative, the IROS 2025 Startup Forum, will facilitate connections between academic research and entrepreneurial ventures, focusing on investment opportunities [2][4] Event Details - The theme of the Startup Forum is "Entrepreneurs and Investors at West Lake," emphasizing collaboration between academia, startups, and capital [6] - The forum will take place during the IROS 2025 conference, with specific timing and location to be announced [6] - Target participants include entrepreneurs in embodied intelligence and robotics, angel investors, venture capital representatives, and local government officials [6] Registration Information - Interested parties can register via a provided QR code or link, with a deadline set for October 12, 2025 [7][9] - Registration requires details about core technologies, funding needs, and project ideas for entrepreneurs, while investors must specify their focus areas [9] Organizer's Value Proposition - The organizers aim to bridge the gap between academic innovation and entrepreneurial initiatives, facilitating efficient resource matching and project implementation [10] - Support will be provided for high-quality participants to foster industry incubation and collaboration in the field of embodied intelligence and robotics [10]

Core Insights - Alibaba's Tongyi Qianwen model is expanding from virtual interactions to physical applications, marking a significant move into embodied intelligence, positioning Alibaba alongside OpenAI and Google in this field [1][13]. Group 1: Model Development and Capabilities - Lin Junyang, head of the Tongyi Qianwen model, has formed a research team focused on robotics and embodied intelligence, indicating a strategic shift towards real-world applications [1][2]. - The Tongyi Qianwen model has achieved full-modal coverage, with its flagship model Qwen3-Max ranking among the top three globally in coding and deep reasoning capabilities, surpassing competitors like GPT-5 [6][12]. - The newly launched Qwen3-VL model enhances capabilities in recognizing small objects and 3D spatial modeling, directly supporting tasks such as dynamic path planning for robots [8]. Group 2: Market Position and Growth - According to a report by Sullivan, Tongyi Qianwen holds a 17.7% market share in China's enterprise-level model invocation market, leading over competitors ByteDance and DeepSeek [11]. - Tongyi Qianwen has open-sourced over 300 full-size models, with a total download exceeding 600 million, facilitating extensive scenario validation for technology implementation [12]. Group 3: Strategic Direction and Investment - The focus on AI physical intelligence is seen as a crucial strategic shift for large model enterprises, with predictions of over $4 trillion in global investments in this area over the next five years [13][14]. - Alibaba has accelerated its investments in embodied intelligence, including a $140 million lead investment in a robotics company, indicating a commitment to deep collaboration in cloud computing and AI platforms [14]. Group 4: Competitive Landscape and Challenges - The competition in the embodied intelligence sector is intensifying, with predictions of global shipments of embodied robots reaching tens of thousands by 2025 and a market size exceeding 232.6 billion yuan by 2030 [19]. - Challenges remain for Alibaba in translating its virtual technology advantages into physical engineering capabilities, particularly regarding domestic component production and cost control [19][20]. Group 5: Future Outlook - The performance of AI foundational models is converging, making the ability to implement in the physical world a new competitive differentiator [20]. - Alibaba aims to establish industry standards in core scenarios such as industrial manufacturing and logistics by leveraging its Tongyi Qianwen model as a brain operating system for robots [20][21].

Core Viewpoint - The strategic partnership between Wuhan Haocun Technology Co., Ltd. and Lingxin Qiaoshou (Beijing) Technology Co., Ltd. aims to integrate Haocun's "Motor Neuron Central System" with Lingxin's advanced dexterous hand technology to overcome technical bottlenecks in humanoid robots' fine operations and intelligent interactions, promoting the large-scale application of embodied intelligence in various sectors such as industry, healthcare, and services [1][6][11]. Technical Integration - The collaboration focuses on deep cooperation in humanoid robot motion control and fine operations, creating a complete technical loop from action decision-making to end execution [1][6]. - Haocun's "Motor Neuron Central System" achieves a precision of 0.001 degrees and a data transmission delay of 3 milliseconds, addressing core challenges in humanoid robot motion control [3]. - Lingxin's Linker Hand series dexterous hands have a freedom degree of 25-30 for the industrial version and up to 42 for the research version, surpassing competitors like the UK's Shadow Hand and Tesla's Optimus Prime [3]. Industry Empowerment - The partnership aims to accelerate the commercialization of humanoid robot technology across diverse scenarios, particularly in industrial manufacturing, healthcare, and domestic services [6]. - Haocun's motion control system combined with Lingxin's dexterous hands will significantly enhance robots' flexibility and precision on assembly lines, with existing collaborations with major industrial players like Foxconn and BYD [6]. - The integration of technologies from both companies represents a crucial pathway for humanoid robots to achieve fine operations, with potential applications in rehabilitation training and high-sensitivity commercial tasks [6][11]. Synergistic Effects - The complementary technologies of both companies create significant synergies, forming a more complete "perception-decision-execution" loop [8]. - Haocun's system acts as the "brain" of the robot, while Lingxin's dexterous hands serve as the "hands," enhancing the robot's capabilities for fine manipulation [8]. - The collaboration aligns with Haocun's strategy to seek partnerships across the industry chain to meet the growing market demand for robotic training [8]. Summary - The partnership between Haocun Technology and Lingxin Qiaoshou reflects a trend in the Chinese robotics industry towards ecological collaboration, aiming to accelerate core technological breakthroughs and industrial applications [11].

Group 1 - ZhiYuan Robotics and Longqi Technology have reached a strategic cooperation with a framework order worth several hundred million yuan for the ZhiYuan Spirit G2 robots, deploying nearly a thousand units primarily in flat panel production lines [3][4] - Yuanli Unlimited signed a strategic cooperation agreement with Shihua Cultural Tourism Group for a single order amounting to 260 million yuan, marking the largest single commercial order for embodied intelligence globally, focusing on "robot + cultural tourism" [6][7] - Shenzhen Deep Venture Capital established an artificial intelligence and embodied robotics industry fund with a total investment of 1.55 billion yuan, aimed at private equity investment and asset management [8] Group 2 - Japan's Murata Manufacturing, Waseda University, and Tmsuk plan to begin trial production of domestically produced humanoid robots in March next year, targeting two prototypes with different focuses on power and sensitivity [9][10] - Tumi's laparoscopic surgical robot has achieved over 15,000 clinical surgeries globally, with commercial orders exceeding 100 units, solidifying its position as a leading domestic brand [12][14]

Core Viewpoint - SoftBank announced the acquisition of ABB's robotics business for $5.4 billion, disrupting ABB's plans for an IPO of the robotics unit and highlighting SoftBank's aggressive investment strategy [1][9]. Group 1: Acquisition Details - The acquisition is expected to be completed in mid-2026, marking a significant shift in the global robotics market and providing opportunities for Chinese high-end robotics manufacturers [1]. - ABB's robotics business generated $2.3 billion in revenue in 2024, accounting for approximately 7% of ABB's total sales, making the acquisition price relatively high [6][16]. Group 2: Challenges and Risks - SoftBank faces challenges in integrating technology and teams post-acquisition, especially given ABB's recent struggles with profitability and declining sales [3][4]. - There is a risk of talent attrition, with studies indicating that 75% of key employees may leave within three years of a major acquisition [4]. - The competitive landscape in the robotics market is intensifying, particularly from Chinese companies, which are rapidly gaining market share [6][7]. Group 3: Strategic Importance - SoftBank views robotics as a critical area for future growth, aiming to integrate ABB's robotics capabilities with its AI investments to create a comprehensive "Physical AI" ecosystem [9][11]. - The acquisition is part of SoftBank's broader strategy to build a complete AI ecosystem, linking robotics with data centers, energy, and chip manufacturing [14][15]. Group 4: Market Dynamics - The global robotics market is projected to grow from approximately $78 billion to $165 billion by the end of 2029, indicating significant potential for investment returns [16]. - The competitive dynamics are shifting, with traditional leaders like ABB losing market share to emerging players, necessitating a reevaluation of strategies across the industry [7][15].

根据机器人大讲堂 &立德智库不完全统计，2025年9月，中国机器人行业融资活动呈现显著回暖态势，共发 生融资事件56起，较8月的39起环比增长超40%，市场活跃度明显提升。融资总额预估达到54亿-78亿元， 较8月28亿-47亿元的区间实现大幅跃升，反映出资本对机器人赛道，尤其是头部项目和前沿技术领域的出手 力度与信心显著增强。 从融资规模来看，大额融资事件表现尤为抢眼。 本月亿元级融资高达 13起，不仅在数量上近乎8月（7起） 的两倍，其战略意义更为深远。其中， 通用具身智能模型研发领域出现单笔高达 10亿元的A+轮融资 ，显示 出市场对这一前沿制高点的强烈追捧； 服务机器人 领域亦有龙头企业完成约 10亿元融资 ，凸显了其在商用 清洁等成熟场景中实现规模化落地的巨大潜力与商业价值。这些大额融资清晰地表明，资本正以前所未有的力 度，向那些具备定义下一代机器人平台潜力的关键技术节点集中。 融资轮次分布延续了早期为主的格局，但中后期融资活力增强 。 种子 /天使轮与A轮系列融资合计占 比达到 86 %，是创新活力的主要来源。与此同时，B轮及以后的中后期融资事件数量有所增加，如昂泰微精等企业 获得持续资本加持，反 ...

在具身智能驱动的机器人操控领域，基于模仿学习的视觉 —运动策略（ Visuomotor Policies ）已成为实现 机器人自主操作的重要技术路径。这类策略通常依赖于视觉观测与本体感知状态（ Proprioceptive States ）的融合范式，通过融合来自摄像头等传感器的外部环境信息和机器人自身的关节角度、末端位置等内部状 态，实现对动作输出的精确控制。该范式虽已在多个任务中取得显著成效，并被广泛应用于各类机器人学习系 统中，却隐含着一个关键缺陷：策略模型对本体感知状态输入的过度依赖，容易导致其在训练轨迹上产生过拟 合，进而削弱其在未见空间配置下的泛化能力。 具体而言，当策略过度依赖如机器人基座坐标、关节角度等内部状态时，其决策过程会逐渐 "固化"于特定的 空间位置与运动轨迹。一旦测试环境中物体的位置、桌面的高度或机器人的起始位姿发生改变——即便视觉场 景在语义上高度相似——策略的表现也可能急剧下降。这种对空间变化的敏感性，严重制约了机器人在动态、 非结构化现实环境中的适用性，也使得每一项新任务的部署都需耗费大量数据与调试成本。 针对上述痛点，千寻智能的研究 团队 在论文 " Do You Need ...

Core Insights - The article discusses the development of micro-robots with insect-level autonomy and flexibility, aimed at performing tasks in complex environments such as disaster recovery and industrial inspections [1] - A research team from Huazhong University of Science and Technology has created a flexible electronic robot (FEbot) that integrates structural adaptability and embedded computing capabilities [1][3] FEBots Design and Modular Architecture - The FEbot utilizes a modular architecture combined with an oscillatory driving mechanism, differing from traditional robots that rely on precise leg trajectory control [3] - The FEbot consists of two main components: a programmable flexible electronic module and a distributed setae array made from superelastic alloys [5] Oscillatory Drive and Asymmetric Friction - The movement of the FEbot is driven by an oscillatory mechanism that relies on asymmetric friction, allowing for net forward displacement during each cycle [7] - The study identified three key states in the movement cycle, with the optimal configuration achieving a peak speed of 109.5 mm/s at a driving frequency of 450 Hz [9] Multi-Modal Movement and Environmental Sensing - FEbots exhibit remarkable multi-modal movement capabilities, allowing for quick reconfiguration based on task requirements [10] - The I-type design, inspired by centipedes, can navigate tight spaces and carry loads up to 5.1 times its weight, while the II-type design showcases omnidirectional movement [11][12] Embedded Artificial Intelligence - The research team incorporated an embodied intelligence framework into the FEbot, enabling real-time intelligent decision-making through hyperdimensional computing [15] - The FEbot demonstrated autonomous behaviors such as danger avoidance and thermal gradient tracking, showcasing its ability to process multiple sensor inputs simultaneously [18]