Let’s Get Physical: The Real Numbers Behind Robotics and Physical AI
3 SEPT, 2026
By Andrew Ye from Global X ETFs
| Marketing Communication. Capital at risk. For Professional Investors Only. Please read fund legal documentation before making any final investment decisions. |
Robots that can perceive their surroundings, interpret instructions and act in the real world are moving beyond controlled demonstrations into factories, hospitals and logistics networks. This shift, often described as physical AI, combines perception, decision-making and physical action. Crucially, it is beginning to produce measurable evidence through installed systems, completed procedures, operating fleets and multi-year commercial commitments.
An established - and changing - market
The foundations are already substantial. Manufacturers installed 542,000 industrial robots in 2024, taking the global operational stock to 4.7 million units. Annual installations have remained above 500,000 for four consecutive years.1 At the same time, collaborative robots, or cobots, are widening the market by working safely alongside people and offering smaller manufacturers a more flexible route into automation. Their share of annual industrial robot installations rose from 2.8% in 2017 to 11.9% in 2024.2
Adoption is also spreading across business models. In healthcare, Intuitive Surgical's da Vinci systems create recurring demand for instruments, accessories and servicing after the initial placement.3 In logistics, Amazon deployed its one-millionth robot in 2025 and introduced AI software designed to improve fleet efficiency.4 Industrial groups such as FANUC are combining established hardware with AI tools that allow operators to issue natural-language instructions, potentially reducing programming time and increasing the value of robots already installed.5
See thee full report: Let’s Get Physical: The Real Numbers Behind Robotics and Physical AI
Humanoids move towards production
Humanoid robots remain earlier-stage, but deployments are becoming more tangible. Figure 02 supported the production of more than 30,000 BMW X3 vehicles during an 11-month deployment, moving over 90,000 components. BMW has since expanded the programme to Figure 03 for logistics-sequencing work.6
The opportunity may extend beyond any single humanoid manufacturer. Google DeepMind is developing Gemini Robotics models intended to transfer learned behaviour across robots of different shapes and sizes, supported by partnerships with Apptronik, Boston Dynamics and Agile Robots.7 Schaeffler's agreement to supply a seven-digit number of robot actuators through 2031 further shows how component suppliers may participate without relying on one platform to dominate.8
Value beyond the robot
Physical AI relies on a broader ecosystem of computing, software, sensors, memory and manufacturing capacity. NVIDIA is developing an end-to-end stack spanning simulation models and edge-computing hardware, while infrastructure initiatives in Japan and Germany aim to provide manufacturers with access to shared AI models and compute.9 These capabilities could shorten development cycles and create recurring revenue opportunities around hardware already deployed.
Physical AI has therefore entered a phase where deployment data can be assessed alongside technological ambition. Industrial installations remain high, cobots are gaining share and AI is improving the flexibility of the installed base. For investors, the theme spans several business models - from robot manufacturers and component suppliers to software, semiconductors and service providers. The Global X Robotics & Artificial Intelligence UCITS ETF (BOTZ LN) is designed to provide exposure to key drivers of the robotics and AI theme across a range of industries and applications.
Sources
1. International Federation of Robotics, September 2025, Global Robot Demand in Factories Doubles Over 10 Years (World Robotics 2025 report).
2. International Federation of Robotics, September 2025, World Robotics 2025 Press Conference Presentation.
3. Intuitive Surgical, July 2026, Intuitive Announces Second Quarter Earnings.
4. Amazon, June 2025, Amazon Deploys Over 1 Million Robots and Launches New AI Foundation Model.
5. FANUC Europe, May 2026, FANUC Collaborates with Google.
6. BMW Group Press, June 2026, BMW Group Advances the Use of Physical AI in Production with Figure 03 Project in Spartanburg.
7. Google DeepMind, December 2025, Introducing Gemini Robotics and Gemini Robotics-ER; Boston Dynamics, January 2026; CNBC, March 2026.
8. Schaeffler, January 2026; Humanoid, May 2026, Humanoid Secures Landmark Deal with Schaeffler to Deploy Thousands of Humanoid Robots.
9. NVIDIA Newsroom, multiple publications, April-July 2026.
