Summary

• NEURA Robotics will take over ACTIVE Shuttle hardware, software, and customer service from Bosch Rexroth from 1 October.
• The autonomous vehicles already move materials inside production and logistics environments.
• Integration with NEURA’s wider robotics software will test whether a shared AI layer can add flexibility without undermining industrial reliability.

Europe’s physical AI market is moving closer to the factory floor as NEURA Robotics takes over an autonomous logistics platform that already works inside industrial plants, bringing installed equipment and customer service obligations into a robotics strategy better known for its ambitions around connected intelligent machines.

NEURA Robotics will acquire Bosch Rexroth’s ACTIVE Shuttle automated guided vehicle business with effect from 1 October. The transaction covers the platform’s hardware and software, while NEURA Mobile Robots will assume responsibility for existing customers and service.

ACTIVE Shuttle moves materials between production and assembly locations, including small load carriers, and is already used by industrial companies rather than remaining a prototype. NEURA plans to retain technologies including ACTIVE Fleet Manager and Bosch Rexroth’s ROKIT navigation software while gradually integrating the equipment with its own AI functions and wider robotics platform.

Chief executive David Reger described the longer term objective as an interoperable robotics environment, saying: “Our goal is an open ecosystem in which mobile robots, industrial robots, and humanoids can work together on a shared infrastructure.” The acquisition gives that ambition an installed mobile logistics layer on which to prove whether common software can deliver practical value.

Industrial customers inherit the integration work

A mobile robot supplying parts to an assembly line operates under conditions that are far less forgiving than most office software. It has to navigate safely around people and equipment, arrive predictably, integrate with plant systems, remain available through production shifts, and continue functioning as software changes around it.

Those requirements make service continuity central to the transaction. Customers that originally bought ACTIVE Shuttle from Bosch Rexroth will judge the change through support response, spare parts, compatibility, uptime, and whether NEURA introduces new functionality without destabilising processes that already work.

Manufacturing environments also contain long lived equipment from many vendors, so the commercial value of a shared robotics layer depends on interoperability rather than simply connecting more products from the same supplier. Controllers, warehouse systems, sensors, industrial networks, safety equipment, and production planning software can remain in service for years or decades, which means newer AI systems normally have to fit around existing infrastructure.

NEURA intends eventually to make navigation and fleet capabilities accessible through its Neuraverse platform, potentially allowing common services to work across different robot types. If that extends cleanly beyond NEURA’s own hardware, manufacturers could reduce some of the integration work that currently accompanies mixed fleets; if it remains closely tied to one ecosystem, the same architecture could create another form of vendor dependence.

Robotics is shifting from machines towards operating systems

Conventional industrial automation has already proved that machines can move goods and perform repetitive tasks reliably when their environment is tightly controlled. Newer AI techniques are intended to make those machines more adaptable, allowing them to recognise changing surroundings, learn or select behaviours more quickly, and coordinate tasks that would otherwise require substantial manual engineering.

Greater flexibility does not remove the industrial requirement for predictability. A language model can produce an imperfect draft that an employee checks before use, whereas a vehicle carrying components through an active factory cannot be given the same tolerance for unexplained behaviour. Perception, AI planning, and adaptive software therefore have to operate inside conventional controls for safety, latency, testing, and change management.

Europe enters that transition with a substantial base of industrial expertise in robotics, drives, machine tools, control systems, logistics automation, and sensors. Combining newer AI software with that installed industrial stack offers one route to commercialisation that does not depend on waiting for humanoid robots to become sufficiently reliable and economical for broad factory deployment.

The ACTIVE Shuttle acquisition also brings recurring operational responsibility that is absent from many robotics demonstrations. Once a machine is part of a production process, the supplier has to maintain it, update it, support users, and manage failures over years, which pushes the economics towards software, services, and platform support as well as hardware sales.

NEURA now has to show that a broader AI architecture can improve an established logistics product without sacrificing the characteristics that made industrial automation useful in the first place. Physical AI will gain credibility through systems that keep factories moving on an ordinary Tuesday, not through machines that perform well only when the demonstration has been prepared in advance.