美國兩黨機械人委員會初步報告警告中美部署差距擴大

The National Security Commission on Robotics for Advanced Manufacturing, a bipartisan initiative of the Special Competitive Studies Project, released its 90-page Preliminary Report on October 13, 2026. The commission’s central finding: the United States leads in the technologies that form the brain of modern robotics, including artificial intelligence, computer vision, autonomy and software, yet lacks the scale to deploy robots, produce robotic systems and manufacture critical components domestically.

The report addresses physical AI, its term for artificial intelligence embodied in machines that sense and act in the real world, and argues that leadership will be determined by productive deployment and resilient production capacity rather than model capability alone. The commission frames the document as a diagnosis rather than a prescription; a Final Report with recommendations is scheduled for March 2027.

A Bipartisan Commission and Its Method

The commission is co-chaired by Senator Ted Budd, a North Carolina Republican, Senator Elissa Slotkin, a Michigan Democrat, and SCSP President and CEO Ylli Bajraktari. Members include executives from Playground Global, Path Robotics, FANUC America, Intrinsic, NVIDIA, Andreessen Horowitz, Boston Dynamics, AMD, General Motors and Mytra; the director of the Johns Hopkins Applied Physics Laboratory; and academics from the Massachusetts Institute of Technology and the University of Michigan. Ayanna Howard stepped down as a commissioner in June 2026 to focus on her new role as president of Spelman College.

The analysis covers robotic systems used in advanced manufacturing, from fixed industrial robots and collaborative systems to autonomous mobile robots and emerging general-purpose platforms. The findings draw on publicly available data, academic and industry research, commission deliberations, briefings and site visits, and engagements with manufacturers, workers, technology developers, systems integrators, researchers, educators and investors. The commission labels them preliminary because robotics technologies, markets and deployment models are evolving rapidly and the available evidence remains uneven in several areas.

The Deployment Gap Behind the Finding

According to International Federation of Robotics figures cited in the report, Chinese factories installed approximately 354,200 industrial robots in 2025, 59% of worldwide installations, compared with approximately 38,400 in the United States. By the end of 2024, Chinese firms had more than two million industrial robots operating in factories, which the report describes as more than five times the U.S. stock of 393,700; that stock represented 8.3% of the global total as of 2024.

The United States ranked eighth worldwide in robot density in 2024, at roughly 307 robots per 10,000 manufacturing employees, a fraction of the Republic of Korea’s 1,220 but above China’s density. U.S. annual installations peaked at 40,000 units in 2022, or 7% of the global total, and the U.S. share fell to 6.3% in 2024; preliminary numbers indicate a rebound to 38,400 units in 2025. Most industrial robots installed in the United States were imported, while Chinese suppliers accounted for 55% of installations in their domestic market in 2025.

Only 12.1% of U.S. manufacturing plants reported using robots in 2021, according to U.S. Census Bureau data cited in the report. China’s Made in China 2025 initiative, launched in 2015, set a goal of 1.8 million operational robots by 2025, and the report states China met it early, in 2024, when its operational stock surpassed two million. The commission cautions that these figures measure industrial robots and do not capture newer systems such as mobile robots and humanoids, and that comparable deployment data for those systems are needed before drawing broader conclusions about physical AI leadership.

Ten Findings Across Four Pillars

The report organizes ten preliminary findings around four pillars: the national framework, the talent pipeline, adoption at scale and the innovation ecosystem. At the center sits what the commission calls a supply-demand paradox: manufacturers wait for robotic systems to become cheaper, more adaptable and better proven, while robotics firms cannot lower costs or improve performance without the deployment volume manufacturers provide. The commission describes this first-mover problem as a market failure unlikely to resolve on its own.

On the national framework, the report finds federal robotics initiatives spread across at least four executive departments and numerous agencies, with no single institution accountable for national-scale outcomes. The National Robotics Initiative coordinated more than $250 million across more than 300 projects before the National Science Foundation ended the program in 2022.

On the talent pipeline, a 2024 Deloitte and Manufacturing Institute study cited in the report projects the industry will need to fill approximately 3.8 million jobs by 2033, with up to 1.9 million potentially going unfilled. The commission finds the employment effects of robotics ambiguous and not predetermined, shaped simultaneously by job displacement, job transformation and new job creation. Apprentices make up only 0.3% of the U.S. workforce as of 2025, the report notes, and traditional multi-year apprenticeship programs struggle to keep pace with rapid technological change.

According to the commission’s release, small and medium-sized manufacturers, about 95% of U.S. manufacturing firms, face capital constraints, knowledge gaps, narrow return-on-investment metrics and cybersecurity and liability uncertainty. The report adds that AI-assisted programming, low-code tools and collaborative robots are expanding the feasibility of automation beyond rigid, high-volume production, while safety frameworks and performance standards struggle to keep pace.

On the innovation ecosystem, the report describes a commercialization valley of death for U.S. robotics startups, whose capital-intensive hardware requirements align poorly with traditional venture capital timelines. China accounted for approximately 60% of global rare-earth mine production, 91% of refined output and 94% of sintered permanent-magnet production in 2024, according to the report, which describes such magnets as necessary components in most motors and actuators used in industrial robots; it adds that in 2025 China implemented export restrictions that disrupted industrial supply chains in the United States and its allies. The operational data U.S. firms produce sits locked inside individual companies, in incompatible formats and behind contracts that bar sharing.

The findings also cite security risks: a 2026 national security determination that added foreign-produced advanced robotic devices to the Federal Communications Commission’s Covered List; a vulnerability discovered in early 2026 in a Chinese company’s robot vacuums that exposed live camera feeds, microphone audio and floor plans from nearly 7,000 devices across 24 countries; and a pre-installed backdoor found in 2025 in a quadruped robot made by Unitree Robotics that enabled surveillance of customers worldwide and unauthorized remote control of the robots.

A National Mission and the March 2027 Final Report

The report concludes that the United States needs a national mission organized around deployment, mobilizing private-sector dynamism, research strength, capital markets, regional manufacturing ecosystems and skilled workers. Such a mission would not copy China’s state-directed model or require domestic production of every component; the strategic question, the commission writes, is where concentrated or adversary-linked dependencies create vulnerability and where domestic or allied capacity is essential.

“A software advantage that cannot be embodied, manufactured, deployed, and sustained at scale will not be sufficient to secure American industrial leadership,” Bajraktari wrote in a letter opening the report.

In the release announcing the report, Budd called the report an important step toward harnessing physical AI, automation and robotics, and said reindustrializing the United States is essential to strengthening critical supply chains, securing leadership in emerging technologies and staying ahead of China. Slotkin said the country needs to bring back and expand critical manufacturing processes and shore up vulnerable supply chains without harming American workers.

The commission says such a mission should be judged by national outcomes rather than the number of programs created or dollars announced, citing measures including the breadth of adoption across industries, regions and smaller manufacturers; the time and cost required to move from assessment to productive operation; the share of pilots that advance to repeat deployments; and domestic and allied capacity in critical components. The Final Report, scheduled for March 2027, will propose recommendations for federal, state and regional leaders; manufacturers and technology companies; educators and workforce-development organizations; workers and labor representatives; investors; and research institutions.