Robotics Startups We Think Will Make It Out of the Lab

Agility’s Digit robots are being tested for moving containers and other repetitive warehouse work. The company ranks first because it is designing deployment, safety and fleet operations together.

Figure moves quickly, has exceptional funding and is placing humanoids into automotive and logistics trials. Its challenge is converting impressive demonstrations into reliable hours of paid work.

Neura builds collaborative and humanoid robots in Germany and reports significant industrial orders. The company’s integrated sensors and manufacturing ambition make it a serious European contender.

Physical Intelligence trains general models intended to transfer across different robot bodies and tasks. If the approach works, its software could become more valuable than any single machine.

Skild AI is building a shared foundation model for robots, supported by large-scale simulation and real-world data. The idea is powerful, though deployment evidence must catch up with the funding.

1X designed NEO for homes, where noise, appearance and safe interaction matter as much as strength. That environment is harder than a controlled factory and potentially much larger.

Generalist AI is trying to teach robots broad manipulation skills rather than a fixed sequence of movements. We want to see whether those policies remain dependable across ordinary industrial variation.

Barcelona-based Theker combines vision, grippers and flexible industrial robots for tasks that change frequently. Its European manufacturing focus gives the company a clear commercial entry point.

Apptronik developed Apollo around factory and warehouse tasks and has partnerships with major industrial companies. The practical body design matters more to us than making a robot look uncannily human.

Unitree manufactures comparatively affordable quadrupeds and humanoids in meaningful volumes. Its speed and pricing have made advanced robot hardware accessible beyond a handful of wealthy labs.