Videos of robots running, dancing and manipulating objects are multiplying, yet their real economic presence remains limited. Global figures now reveal the gap between spectacular demonstrations and industrial deployment.

A market can finally be measured

On 30 September 2026, IFR reported around 7,000 full-size humanoids taller than 140 centimetres sold worldwide in 2025. The same report counts almost 250,000 professional service robots in 2025, up 24%. The figure of 542,000 industrial robot installations refers to 2024: the categories and years differ.

These volumes describe a market that remains limited. Industrial, service and humanoid robot categories should not be added together without checking their statistical scope.

Many are still test platforms

A large share are used by laboratories and companies to develop software, collect data and evaluate behaviour rather than perform regular work.

Car makers are interested because a humanoid could use human aisles, tools and workstations without rebuilding a factory.

Very practical obstacles

Battery life, reliable hands, stability, speed and reactions to unexpected events remain major problems. A prepared demonstration is not an eight-hour shift.

Safety is equally central: a powerful machine near people must limit force and remain predictable when something fails.

Work will change gradually

Near-term uses will mostly be pilots for light handling, line supply and repetitive or hazardous work.

They will also create supervision, maintenance and training roles; employment effects will depend on work organisation as much as technology.

Conclusion

The humanoid market exists but remains embryonic. Seven thousand sales are a starting point, not proof of maturity.

The real change begins when robots perform useful work reliably, safely and at a competitive cost.