944 manufacturers, 80 percent of them SMEs: why nobody dominates the service robot market

The number 944 comes from the industry structure chapter of World Robotics 2025 – Service Robots, whose executive summary the IFR makes freely available. The statistical department keeps a list of the service robot suppliers known to it, and two exclusions decide who is on it: 'This excludes prototyping services and system integrators.' Whoever develops robots for others without selling a product of their own does not count. And whoever, like werob, introduces other manufacturers' robots into operations does not count either. What is counted is the producer.

The sales figures in the same volume, by contrast, rest on a sample: the 2025 report is based on returns from 294 suppliers. That is less than a third of the known producers, and the IFR itself says why the gap is large: 'Many companies are still in the funding or prototyping stage and intend to offer a marketable product in the future.' So part of the 944 sell nothing yet. For an operator looking for a machine for next quarter the field is therefore smaller than the number, but it is still far larger than any list a single vendor or integrator knows.

Two structural statements by the IFR describe the field more precisely than the headcount. The first: 'On a global scale, 80% of the companies are small or medium-sized enterprises (SMEs) with up to 500 employees.' Four in five manufacturers are therefore companies for which a single major customer, a missed funding round or a supply bottleneck for one component touches their continued existence. The second: 'Although the service robotics industry is a young and growing industry, 93% of the suppliers are considered incumbents.' Incumbent, in the IFR's usage, does not mean large but established in the market for years, and the group explicitly includes companies from other industries that have added service robots to their range, such as makers of cleaning machines or agricultural equipment.

Then the trend: 'The 2010s saw a wave of new service robot manufacturers. Since then, the number of newly established companies steeply declined.' The IFR gives two reasons. First, some market segments have reached a maturity in which the existing companies grow; autonomous mobile robots for warehouse logistics are named explicitly. Second, founding activity has moved away from hardware: 'Founding activities shifted away from the development of robot hardware', towards software and artificial intelligence. Whoever buys a service robot today most likely buys it from a company that already existed before 2020, employs fewer than 500 people, and whose segment is now changing not through new entrants but through acquisitions; the IFR speaks of 'a lot of merger and acquisition activity'.

The IFR's explanation for the heterogeneity is so simple it is easily read past: 'Service robots for professional use are extremely diverse since they are usually designed to perform a specific task.' An industrial robot is an arm that gets programmed; the same machine welds, paints or palletises. A service robot is built for one task: scrubbing floors, carrying trays, hauling pallets outdoors, patrolling fences, pushing feed. Each of these tasks has its own market, its own customers and its own suppliers, and a manufacturer that leads in one is not even present in the next. The market, in the IFR's words, consists of 'many niche products for professional services and a few high-volume applications'.

The 2024 sales figures show how unequal the niches are. Worldwide, more than 199,000 professional service robots were sold, 9 percent more than the year before. More than every second one was built for transporting goods; the transportation and logistics segment grew 14 percent, and the IFR knows 333 suppliers in this area alone. Hospitality reached the second-largest volume with over 42,000 units sold, even though sales fell 11 percent. Professional cleaning grew 34 percent to over 25,000 units, agriculture contracted 6 percent to close to 19,500, security robots reached 3,128 units, inspection and maintenance close to 2,800. A supplier selling 5,000 robots a year into hospitality is a big name there and an unknown in cleaning. That is why there is no market leader for service robots, at most one per task, and even that only for a year.

One sentence from the IFR deserves particular attention in procurement: 'Many service applications are based on collaborative industrial robots, purchased from an industrial robot producer. The service robot supplier is therefore not considered a robot producer as the robot is purchased from a third party. These companies act like system integrators, combining different components and developing software to create a solution.' Some of the companies that present themselves as robot manufacturers do not build robots. They buy an arm or a mobile platform, add grippers, sensors and software, and sell the result under their own name.

For the operator that is no defect, but it is a question that belongs before the contract: who builds the platform this solution stands on, and what happens to spare parts, software updates and warranty if the supplier and the platform maker part ways? A supplier with 30 employees whose robot runs on another company's platform carries two failure risks, not one. The IFR count makes this layer visible by excluding it: whoever is not counted as a producer does not produce.

The number 944 is not an argument for a particular brand and not one for waiting. It is an argument for a procedure. First: selection starts from the task, not from the manufacturer. Whoever is looking for wet cleaning of circulation areas looks among the suppliers of professional cleaning, and the names from logistics or hospitality are irrelevant to that, however well known. Second: the supplier's size is a property of the product. With 80 percent SMEs, the question of financing, spare-parts stocking and where the software goes in an insolvency is not a courtesy but the check a datasheet does not replace. Third: the origin of the platform belongs in the contract, for the reason above.

And fourth, a yardstick for any list put in front of you: a selection of forty manufacturers, as an integrator or dealer typically carries, covers fewer than one in twenty of the 944 producers. That is not a shortcoming, because an operation does not need 944 options but the three that do its task and have service in its country. But it means a list is always a selection someone has made, and the question about the criteria of that selection belongs at the start of the conversation, not the end.

All figures come from the executive summary of World Robotics 2025 – Service Robots by the International Federation of Robotics and the accompanying press release of 7 October 2025; quotations are given in the English original. The number 944 describes the producers known to the IFR worldwide, not the products available in Europe, and it includes companies without a marketable product. The sales figures rest on returns from 294 suppliers and are minimum values, not a full census. The executive summary publishes no market shares for individual manufacturers; the statement that no manufacturer dominates follows from the structure of the field, not from a share calculation. The full supplier list sits in the paid volume, which this article has not consulted.