In July 2025, Amazon reached a number that hardly made any headlines. The company was for the first time operating more than one million robots in its warehouses worldwide. About three quarters of all orders were already being processed with robot support. Most business media did not report it. When they did, it was as a footnote. That is remarkable — because the million marks a very specific threshold: it is the moment when robotics stops being an experiment and begins to become an infrastructure.
Anyone who wants to understand what is going to happen economically in the next ten years cannot get past this number. Here are seven data points that complete the picture.
The global robotics market is projected to grow from around 90 billion US dollars in 2024 to over 205 billion US dollars by 2030. That is an average annual growth rate of around 15 percent.
For comparison: the global market for artificial intelligence grew at similar rates in the first wave from ChatGPT between 2022 and 2024. Both markets are intertwined — AI is the brain, robotics the hands — and both are at the start of a long curve.
Important here: the 205 billion is a conservative estimate from GlobalData in 2025. More aggressive forecasts, such as those from The Business Research Company, go as high as 365 billion by 2030. Which figure is closer to reality will become clear over time. But even the conservative estimate means a doubling of market size in less than six years.
Data point 2 — The humanoids are coming
Within the robotics market, one category is growing significantly faster than the rest: humanoid robots. Goldman Sachs forecasts the humanoid robot market to grow ninefold by 2030.
The reason is mechanically simple. Industrial robots only work where infrastructure has been built for them — movable rails, defined work stations, clearly demarcated spaces. Humanoid robots are designed to work where people work — in the same rooms, with the same tools, without infrastructure having to be rebuilt.
Tesla is building Optimus. Figure AI unveiled the third generation of its robot in October 2025. Boston Dynamics is bringing the next generation of its famous platform with the electric Atlas. 1X Technologies from Norway launched Neo, the first humanoid robot for household use. As of 2026, all four companies are still in the early phase of delivery — but the pipeline is full.
Data point 3 — The labour shortage
In Europe, North America and parts of Asia, there is a shortage of workers for physical and repetitive tasks. This is not a political assertion, but a measurable reality. In Germany, according to the Federal Employment Agency, almost 700,000 skilled workers were missing in 2025. In Switzerland, the economy has been talking for years about a chronic skilled-worker shortage in the service and care sectors.
The reasons are structural. The baby boomers are retiring. Younger generations are smaller in number. More people are moving into digital and academic professions. Practical training is losing applicants.
For companies this means: wage costs for physical work continue to rise while the availability of workers falls. By pure market logic, robotics thus becomes an economically rational solution — not out of fascination, but out of business necessity.
Data point 4 — Unit costs
What made robotics a specialist project ten years ago was the price. An industrial robot cost six-figure sums, a humanoid robot simply was not for sale.
Today this looks different. 1X Neo is offered from around 20,000 US dollars. Figure AI designed its BotQ production facility for 12,000 units per year, with the goal of further reducing unit costs. With Optimus, Tesla is applying its core competency of mass production — and although today’s prices are not yet publicly communicated, the pipeline leaves no doubt about the direction.
With each year, unit costs fall further. What was a specialist project in 2020 is a commodity product with a clear price tag in 2026. By 2030, the same argument will apply to industries that hardly come into contact with robotics today.
Data point 5 — Industry penetration
Anyone who thinks robotics only concerns warehouses and production is missing the breadth of today’s application. Pudu Robotics from China — one of the largest providers in the service robotics segment — had delivered over 100,000 robots worldwide by November 2025. The deployment areas range from restaurants and hotels to hospitals and cleaning companies.
Hikrobot specialises in warehouse automation and delivers high-rack robots worldwide. Mushiny from China has equipped hundreds of e-commerce warehouses worldwide with its goods-to-person system. Realman, also from China, has developed a robotics massage robot that is already being deployed in spas and wellness centres.
These are not the names that appear in Western business media. But they are the companies that already have hundreds of thousands of robots in operation today. The wave is not in the future. It is the present.
Data point 6 — The asset class
Until now, robotics was a hard-to-access asset class for private investors. Anyone wanting to invest in robotics could buy shares in robotics manufacturers — usually with a leverage that covered only a fraction of the sector. Direct participations in robot fleets were reserved for institutional investors with large tickets.
This is changing. The tokenisation of real assets, often referred to as Real World Assets, has established itself as a serious investment category over the past two years. Switzerland has gained world-leading status with the DLT Act of 2021. Companies such as Sygnum, Aktionariat and Beep Labs use this foundation to make investments that were previously accessible only to large investors more broadly available.
For robotics this means: for the first time, private investors can participate in an asset category that was previously distributed within corporations.
Data point 7 — The missing headline
Back to the beginning. One million robots at Amazon — and hardly anyone is talking about it. This is not an oversight of the business press. It is the same pattern we have seen in other technology waves.
ChatGPT reached one hundred million users in two months in November 2022. Only six months later did the first serious articles appear on what this means economically. The iPhone was introduced in June 2007. Only in 2010, three years later, did the broad business press begin to understand that it was not about a phone, but about the platform of the next decade.
Robotics is at a similar point right now. The million at Amazon is not the end of a development — it is the beginning. The headlines come later. And with them attention, capital, valuations.
Anyone who is already looking today is not too early. They are exactly in time, where the next great economic wave begins.
