Autonomous Robots

Fanuc’s May partnership with Google to deploy factory robots controlled by natural language commands signals Japan’s strategic pivot toward physical AI, a shift that could reshape the industrial robotics market worth billions annually. For institutional investors, this represents a rare convergence of Japan’s entrenched hardware dominance with AI infrastructure owned by U.S. tech giants, creating both opportunity and dependency risks in a sector where Chinese competitors are rapidly gaining share.

• Fanuc partnered with Google on May 13 to build factory robots controlled by spoken and handwritten instructions without programming skills required.
• Japanese manufacturers held 80 percent of the global industrial robot market in the 1990s, but that share has fallen to roughly 40 percent today.
• China’s humanoid robot makers now account for 40 percent of that market as of 2024, though many still depend on Japanese machinery components.

• 40% Japan’s current share of global industrial robot market, down from 80 percent in the 1990s
• 40% Chinese companies’ share of the global humanoid robot market as of 2024
• $4.9B Japan’s annual digital services deficit recorded by the Ministry of Economy, Trade and Industry

Japan is reasserting itself in artificial intelligence not through competing with OpenAI or scaling large language models, but by embedding AI into factory robots that can perform complex, unstructured tasks without human reprogramming.

Fanuc, the world’s largest industrial robotics manufacturer founded in 1956, announced on May 13 that it will develop AI-powered factory robots jointly with Google that respond to spoken and handwritten instructions. The system, built on Google Gemini, eliminates the need for specialized programming knowledge to operate robots on production lines.

The announcement marks a deliberate strategy by Japanese manufacturers to leverage decades of accumulated factory-floor data and motion-control expertise, knowledge that was historically passed down tacitly among workers, and convert it into machine learning training material for what the industry calls physical AI.

Fanuc opens proprietary robot software after decades of closed architecture

The Fanuc-Google partnership represents a watershed moment for Japanese manufacturing culture. For more than three decades, Fanuc maintained closed, proprietary robot software systems that locked customers into its hardware ecosystem. That approach prioritized control and hardware margins over interoperability.

But on May 13, Senior Managing Officer Kenishiro Abe acknowledged at a press conference that Fanuc cannot build a competitive AI ecosystem alone and must now integrate software from multiple vendors.

The partnership stems from the limitations of developing an entire AI ecosystem in-house. It plans to incorporate AI systems from a host of different companies.

Kenishiro Abe, Senior Managing Officer, Fanuc

In December 2024, Fanuc had already signaled this shift by announcing a collaboration with NVIDIA to open its previously closed robot software infrastructure.

The dual partnerships with Google and NVIDIA, both American companies, reflect a calculated dependency: Fanuc retains manufacturing and robotics expertise while outsourcing the AI development layer to firms with superior training infrastructure and model sophistication. This reverses the traditional Japanese manufacturing playbook of vertical integration and proprietary systems.

For Fanuc’s customers, the change means robots can be configured and adapted without hiring specialist programmers, lowering barriers to automation deployment across mid-sized manufacturers globally.

Chinese robot makers gain 40 percent market share while relying on Japanese components

China’s rise in humanoid robotics is real but incomplete. As of 2024, Chinese companies such as Estun Automation and Inovance Technology control approximately 40 percent of the global humanoid robot market, matching Japan’s current share of industrial robotics overall. This represents a dramatic acceleration, just years ago, Chinese manufacturers held single-digit shares.

However, the Nomura Securities analysis reveals a structural dependency: many Chinese robot makers still source critical components from Japanese suppliers, particularly precision manipulators (robot hands), motion-control systems, and semiconductor equipment manufacturing tools.

Japan’s manufacturing base still commands distinct advantages that cannot be rapidly replicated. The International Federation of Robotics data shows that in the 1990s, Japanese companies accounted for 80 percent of global industrial robot production.

The subsequent decline to 40 percent reflects market share loss primarily to Chinese competitors and South Korean firms, not technological obsolescence. What Nomura Securities identifies as Japan’s enduring edge is factory-floor data accumulated over decades, combined with expertise in precision motion control and the institutional knowledge embedded in Japanese supply chains.

Chinese competitors can manufacture robots at competitive prices, but lack the decades of manufacturing datasets and motion-tuning expertise that Fanuc, ABB, and Yaskawa possess.

Japan confronts $4.9 billion annual digital services deficit while chasing physical AI leadership

Japan faces a paradox: it dominates hardware and robotics but imports most of its digital software infrastructure from the United States. The Ministry of Economy, Trade and Industry recorded a $4.9 billion annual digital services deficit, money flowing out of Japan to pay for software, cloud services, and AI models developed elsewhere.

This structural imbalance explains why Fanuc and other Japanese industrial leaders are forced to partner with Google and NVIDIA rather than building AI capabilities independently.

The deficit reflects decades of Japanese industry prioritizing manufacturing excellence over software product development.

While Japanese companies excel at embedded systems, precision control, and industrial automation, they lack the venture capital ecosystem, immigration policies for tech talent, and cultural emphasis on software startups that enabled American and Chinese firms to dominate AI training and inference infrastructure.

The Fanuc-Google partnership partially addresses this gap by giving Japanese roboticists access to Gemini’s training and reasoning capabilities, but at the cost of vendor lock-in and continued dependence on foreign AI providers.

Nomura Securities projects that a post-2030 industrial economy will reward the convergence of motion-control expertise, industrial datasets, and precision manipulation technology, precisely where Japan holds competitive advantage.

However, realizing that advantage requires Japan to simultaneously close its digital deficit and maintain component-supply dominance as Chinese competitors accelerate up the value chain.

Fanuc’s compatibility pledge signals bet on Google ecosystem dominance in factory AI

Fanuc announced plans to make all its robots compatible with Google software, a forward-looking commitment that ties the company’s future growth to Google’s success in enterprise AI markets. This represents a strategic choice: rather than betting on an open-source or vendor-neutral robot operating system, Fanuc is endorsing the Google stack.

The decision implies confidence that Google’s Gemini models and infrastructure will outcompete alternatives, and carries the implicit risk that if Google’s robotics AI offering stalls or fragments, Fanuc’s installed base becomes vulnerable.

Fanuc’s full compatibility pledge with Google software remains conditional on Google sustaining AI development velocity in industrial applications, a domain where results remain unproven at scale.

For institutional investors evaluating robotics and industrial automation exposure, the Fanuc-Google partnership clarifies competitive positioning but introduces execution risk.

Japanese manufacturers possess irreplaceable datasets and motion-control talent; American cloud and AI providers control model training and inference infrastructure; and Chinese competitors are willing to operate at lower margins to gain market share and component data.

The outcome will depend on whether Japanese factories can be retrofitted with effective physical AI systems before Chinese or American competitors build their own supply chains or develop indigenous robotics expertise.

Watch for Fanuc’s first production deployment milestones with Google Gemini-powered robots in 2025 and 2026, as well as any announcements from NVIDIA on how its December 2024 software collaboration with Fanuc translates into commercially available robot control platforms. Equally important: monitor whether the METI’s digital services deficit narrows or widens as Japanese manufacturers increase their dependence on Google and NVIDIA infrastructure, a metric that will signal whether Japan is closing the AI gap or deepening its structural dependence on foreign software vendors.