Group 1 - In 2025, Shenzhen's total import and export volume reached 4.55 trillion RMB, a year-on-year increase of 1.4%, marking a historical high and maintaining its position as the leading city in mainland China for foreign trade [1] - Shenzhen's import and export scale has maintained positive growth for six consecutive years, accounting for 10% of the national total and 48% of Guangdong's total, contributing 15.5% to the province's growth [1] - The export scale reached 2.74 trillion RMB, while imports grew by 8% to 1.81 trillion RMB, contributing 53.2% to the province's import growth [1] Group 2 - There were 62,300 enterprises engaged in import and export activities in Shenzhen, a 16% increase year-on-year, with over 8,600 new enterprises added [3] - The top 100 enterprises in Shenzhen achieved an import and export scale of 2.5 trillion RMB, growing by 13.6%, significantly driving overall growth [3] Group 3 - Private enterprises are the main force in foreign trade, with 55,100 private enterprises generating an import and export value of 3.12 trillion RMB, accounting for nearly 70% of the city's total [4] - Shenzhen's high-tech product import and export scale reached 1.4 trillion RMB, ranking first in the country and accounting for 30.7% of the city's total [5] Group 4 - Shenzhen's exports of modern manufacturing products, including digital cameras and medical devices, ranked first in the country, totaling 109.77 billion RMB, with a growth of 17.4% [5] - The AEO (Authorized Economic Operator) system has become a "pass" for many enterprises, with 181 AEO enterprises benefiting from approximately 100 million RMB in fee reductions in 2025 [6] Group 5 - In 2025, Shenzhen's imports and exports to countries involved in the Belt and Road Initiative reached 1.571 trillion RMB, accounting for 34.5% of the city's total [8] - The international logistics network has been enhanced, with the number of international and regional air routes from Shenzhen Airport increasing to 43, covering 27 countries and regions [8] Group 6 - Shenzhen's cross-border e-commerce overseas warehouse export scale ranks first in the country, with simplified customs declaration processes improving efficiency by nearly one-third [10] - The city is poised to continue its high-level opening and high-quality development in the face of complex external environments, leveraging its optimized structure and resilient industries [10]

Core Viewpoint - The research conducted by a collaboration of institutions including Shanghai Jiao Tong University and Bosch has developed a new framework called Mantis, which enhances robot learning by enabling them to predict future actions, significantly improving their efficiency and performance in complex tasks [3][4][8]. Group 1: Mantis Framework - Mantis introduces a "decoupled visual foresight" capability, separating the tasks of "seeing the future" and "performing actions" to optimize each task's effectiveness [4][9]. - The framework employs a technique called "latent action queries," which helps robots identify necessary actions from current and future scenes, enhancing their learning process [4][6]. Group 2: Training Methodology - Mantis utilizes a progressive training approach, starting with observing human operation videos, then integrating real robot operation data, and finally incorporating language understanding training [5][6]. - This method allows robots to learn from simpler tasks before advancing to more complex ones, akin to how children learn [5]. Group 3: Performance Metrics - In tests on the LIBERO simulation platform, Mantis achieved a success rate of 96.7%, outperforming several advanced systems like OpenVLA and π0 [6][9]. - Mantis demonstrated a significantly faster learning speed, achieving results in a few training cycles compared to traditional methods that require many more [6][9]. Group 4: Real-World Testing - The research team conducted real-world tests across three scenarios to validate Mantis's capabilities, showing superior performance in understanding world knowledge, basic reasoning, and intent comprehension [7]. - Mantis exhibited strong generalization abilities, particularly in handling novel instructions, outperforming the leading open-source model π0.5 [7]. Group 5: Future Implications - The development of Mantis signifies a new direction in robotics, balancing operational skills with language comprehension, which is crucial for future integration into human life [8]. - The potential applications of robots with visual foresight span various sectors, including household chores, healthcare, manufacturing, and service industries, promising more precise and efficient assistance [8].

Core Viewpoint - The article argues that China is not "in trouble" as some in the U.S. claim, and the U.S. should abandon the notion of "defeating China" [1] Group 1: Economic and Trade Relations - The article highlights the U.S.'s initial confidence in defeating China in trade, which has been proven to be misplaced as China has demonstrated its strength in this area [1] - Despite the GDP gap, China leads significantly in critical future-oriented sectors such as rare earth production, solar capacity, electric vehicles, and industrial robotics [2] - A recent report from the U.S.-China Economic and Security Review Commission states that China is at the forefront of quantum communication and making rapid advancements in quantum computing and sensing [2] Group 2: Technological Advancements - China is making significant strides in chip technology, with reports indicating efforts to master lithography technology [2] - The article emphasizes China's advancements in electric vehicles, clean energy, and robotics, which are crucial for future competitiveness [2] Group 3: Military Capabilities - While the U.S. maintains the strongest military globally, China is rapidly catching up, particularly in naval capabilities with its latest aircraft carriers [3] - The article acknowledges China's economic and demographic challenges but notes that the government's "14th Five-Year Plan" is addressing these issues [3] Group 4: Soft Power and Perception - China's soft power is gaining traction in the U.S., as evidenced by the popularity of platforms like TikTok and the positive experiences of American tourists in China [3] - The article suggests that Americans are beginning to reassess China's global standing, indicating a shift in perception [3] - The article has sparked positive reactions among American netizens, advocating for more cooperation with China, while also drawing attention from Japanese netizens who criticize their own right-wing views [3]

Core Insights - The article highlights the rapid advancements in technology expected by 2026, including emotional AI, wearable health devices, renewable energy innovations, and more, which are set to transform daily life and work environments. Group 1: Emotional AI - The next generation of emotional AI assistants is surpassing traditional ones like Siri and Alexa by understanding human emotions through tone, expression, and wording, providing personalized relaxation recommendations and enhancing customer care [2] - Emotional AI is becoming a partner in daily life, playing a significant role in mental health support and personal growth [2] Group 2: Wearable Devices - By 2026, wearable devices are expected to evolve into "life guardians," with smart skin patches monitoring hydration, stress hormones, and nutrition levels, while AI watches can predict heart disease risks days in advance [3] - These innovations are shifting healthcare from reactive treatment to proactive prevention [3] Group 3: Renewable Energy - New night-time solar panels will capture infrared radiation, allowing for continuous power supply even in darkness, while advancements in lithium-sulfur and solid-state battery technologies will enhance energy storage for electric vehicles [4] - 2026 may mark a historical turning point for the decline of fossil fuels as the dominant energy source [5] Group 4: Brain-Computer Interfaces - Brain-computer interface technology is moving from medical applications to everyday use, enabling paralyzed patients to control devices with their thoughts and enhancing work efficiency through instant data retrieval [6] - By 2026, these interfaces will be integrated into daily life, impacting prosthetics, rehabilitation, and augmented reality experiences [6] Group 5: Space Economy - The space economy is set to flourish, with suborbital flight prices dropping below $100,000 and companies developing zero-gravity hotels, indicating a new era of space exploration and commercialization [7] Group 6: AI in Education - AI-driven personalized education is emerging, adapting courses to individual learning styles and helping adults with career transitions, fundamentally reshaping educational approaches [9] Group 7: Quantum Computing - Quantum computing is expected to solve complex problems beyond the capabilities of classical computers, significantly reducing drug development timelines and optimizing supply chains [10] Group 8: Smart Cities - Cities will become smarter by 2026, with AI traffic systems alleviating congestion, IoT-enabled waste management, and predictive security systems enhancing urban living [11] Group 9: Robotics - Robotics will become ubiquitous in households and various work environments, with cooking robots, construction robots, and personal care robots transforming daily tasks [12][13] Group 10: Bioprinting - The bioprinting field is anticipated to achieve significant advancements, enabling the creation of disease models, personalized medicine, and potentially addressing organ donation shortages [14]

Financial Performance - In the third quarter of 2025, the total revenue was RMB 407 million, a slight decrease of 0.2% compared to RMB 408 million in the same period of 2024 [1] - The net loss for the third quarter of 2025 was RMB 100 million, an increase of 20.2% from a net loss of RMB 83.2 million in the same period of 2024 [1] Strategic Direction - 2025 is a critical year for RoboSense as it expands from lidar technology to a robotics technology platform [1] - The company has a clear path: the automotive business provides scale advantages and a technological foundation, while the robotics business contributes high growth [1] - The Robotaxi business opens up long-term growth potential, with a focus on chips, algorithms, and digital architecture to become a "global leading robotics technology platform company" [1]

Core Viewpoint - The recovery momentum of Chinese tech stocks is still in its early stages as China emerges as an AI superpower, attracting Western capital back to the market [1] Group 1: Investor Sentiment - Investors are eager to position themselves as their portfolios lack exposure to the Chinese AI trend [2] - Global investors have begun reallocating funds to some of China's largest and most liquid stocks, benefiting from recent policy stimulus and technological breakthroughs [3] Group 2: Market Comparisons - The recovery of Chinese tech stocks is at an earlier stage compared to the Nasdaq's growth, with lower price-to-earnings ratios compared to similar growth and profitability peers in the U.S. [4] Group 3: Foreign Investment Interest - U.S. investment firms managing global funds are increasingly interested in Chinese stocks, with investors from the Middle East, Southeast Asia, and Europe focusing on gaining exposure to China [5] - Middle Eastern investors seek stable long-term capital in China's digital economy, while European investors are drawn to the Chinese market due to a lack of local AI firms [6] Group 4: IPO Activity - The number of Chinese companies applying for IPOs in Hong Kong has reached a high level, covering various sectors including internet, software, AI, robotics, and tech supply chains [6] - Hong Kong IPO fundraising reached HKD 216 billion (approximately USD 27.8 billion) in the first ten months of this year, more than three times the amount from the same period last year [7] Group 5: Positive Outlook from Foreign Investment Banks - Despite a global tech stock pullback, several foreign investment banks are bullish on the Chinese stock market, citing advancements in the tech sector as a key reason [8] - Morgan Stanley predicts further gains in the Chinese stock market by 2026, with target prices for MSCI China Index, Hang Seng Index, and CSI 300 Index set at 90 points, 27,500 points, and 4,840 points respectively [9] - UBS anticipates another prosperous year for the Chinese stock market, driven by favorable factors including innovation, with a target for the MSCI China Index at 100 points, indicating a 14% upside from current levels [10]

Core Insights - The article discusses the development of RoboTron-Mani, a universal robotic operation strategy that overcomes the limitations of existing models by integrating 3D perception and multi-modal fusion, enabling cross-platform and cross-scenario operations [1][3][21]. Group 1: Challenges in Robotic Operations - Current robotic operation solutions face a "dual bottleneck": either lacking 3D perception capabilities or suffering from data set issues that hinder cross-platform training [2][3]. - Traditional multi-modal models focus on 2D image understanding, which limits their ability to interact accurately with the physical world [2][3]. - Single data set training leads to weak generalization, requiring retraining for different robots or scenarios, which increases data collection costs [2][3]. Group 2: RoboTron-Mani and RoboData - RoboTron-Mani is designed to address the challenges of 3D perception and data modality issues, achieving full-link optimization from data to model [3][21]. - The architecture of RoboTron-Mani includes a visual encoder, 3D perception adapter, feature fusion decoder, and multi-modal decoder, allowing it to process various input types and produce multi-modal outputs [5][7][9][10]. - RoboData integrates nine mainstream public datasets, containing 70,000 task sequences and 7 million samples, addressing key pain points of traditional datasets by completing missing modalities and aligning spatial and action representations [11][12][15][16]. Group 3: Experimental Results and Performance - RoboTron-Mani has demonstrated superior performance across multiple datasets, achieving a success rate of 91.7% on the LIBERO dataset, surpassing the best expert model [18][21]. - The model shows an average improvement of 14.8%-19.6% in success rates compared to the general model RoboFlamingo across four simulated datasets [18][21]. - Ablation studies confirm the necessity of key components, with the absence of the 3D perception adapter significantly reducing success rates [19][22]. Group 4: Future Directions - Future enhancements may include the integration of additional modalities such as touch and force feedback to improve adaptability in complex scenarios [23]. - There is potential for optimizing model efficiency, as the current 4 billion parameter model requires 50 hours of training [23]. - Expanding real-world data integration will help reduce the domain transfer gap from simulation to real-world applications [23].

点击下方 卡片 ，关注" 具身智能 之心 "公众号 作者丨 Jiahong Chen等 编辑丨具身智能之心 本文只做学术分享，如有侵权，联系删文 >> 点击进入→ 具身智能之心 技术交流群 更多干货，欢迎加入国内首个具身智能全栈学习社区 ： 具身智能之心知识星球 (戳我) ， 这里包含所有你想要的。 在机器人操控领域，" 通用化 " 与 " 轻量化 " 的矛盾长期制约着技术落地——现有视觉-语言-动作（VLA）模型虽能实现复杂任务推理，但因参数量庞大、计算需求 高，难以部署在移动机器人、嵌入式系统（如 Jetson Orin Nano）等资源受限设备上。 而由英属哥伦比亚大学、阿尔伯塔大学与小米汽车团队联合提出的 NanoVLA ，用 " 视觉-语言解耦融合+长短动作分块+动态路由 " 的创新架构，彻底打破这一困 境：既保留通用 VLA 模型的任务精度与泛化能力，又将推理速度提升 52 倍、参数量压缩 98%，首次实现 "在边缘设备上高效运行通用机器人策略" 的目标。 为什么要重构 VLA 模型的边缘部署逻辑？ 当前主流 VLA 模型陷入 "性能与效率不可兼得" 的困境：为实现跨任务泛化，模型通常依赖数十亿参 ...

Core Viewpoint - The article discusses the contradictions in the Trump administration's manufacturing repatriation policies, highlighting how tariffs and immigration policies are counterproductive to the goal of bringing manufacturing jobs back to the U.S. [1][2][4] Group 1: Policy Contradictions - Tariffs have increased the costs of raw materials and imported equipment for U.S. manufacturers, making it harder for them to expand [1][7] - New immigration policies may reduce the supply of skilled labor, exacerbating the shortage of workers in the manufacturing sector [1][10] - The administration's budget cuts threaten the subsidies necessary for companies to repatriate jobs, leading to an unstable subsidy environment [1][14] Group 2: Business Reactions - Business leaders express frustration over the lack of clarity in the "America First" policies, making it difficult for them to plan investments [1][7] - Companies like Ford have reported significant cost pressures due to tariffs, which hinder their ability to invest more in the U.S. [7][8] - The uncertainty surrounding tariffs and their potential continuation after the Trump administration creates a paralysis effect, preventing companies from approving repatriation projects [6][7] Group 3: Labor Market Challenges - There are currently over 400,000 manufacturing job vacancies in the U.S., particularly in high-tech fields like semiconductors and robotics [9] - New immigration policies, including increased fees for H-1B visas, may further complicate the ability of companies to hire skilled foreign workers [10][13] - Industry groups warn that the changes in immigration policy could undermine the talent pipeline necessary for rebuilding the manufacturing sector [12][13] Group 4: Funding and Investment Issues - The Trump administration has touted trillions in investment commitments from companies, but internal budget cuts have created uncertainty regarding subsidies that could stimulate further investment [5][14] - The White House has ordered a review of federal funding and loans, injecting additional uncertainty into the investment landscape [14] - The lack of a coherent strategy for funding and tariffs leaves companies confused about how to proceed with their manufacturing plans [14]

Core Viewpoint - Dongtu Technology (300353) does not directly produce humanoid robots but focuses on providing foundational technology support in the field of embodied robots [1] Group 1: Company Overview - The company's Hongdao operating system is involved in national major projects and possesses fully autonomous and controllable attributes [1] - Dongtu Technology's AUTBUS bus network is the only domestic bus technology that has received international standards certification [1] Group 2: Technological Capabilities - The combination of the Hongdao operating system and AUTBUS bus network can create a robotic electronic architecture with time-deterministic networks and deterministic computing capabilities [1] - This architecture supports the high safety, high performance, and high-quality development of embodied robots, providing autonomous and controllable foundational technology assurance [1]