NEURA Robotics and SECO anchor a European humanoid compute supply chain on Qualcomm silicon
NEURA Robotics has entered a strategic partnership with the Italian embedded-computing specialist SECO, which will design, engineer, and manufacture the electronic compute modules used inside NEURA's cognitive robots, including the 4NE1 humanoid. The announcement came on September 7, 2026, with Qualcomm Dragonwing processors at the center of the hardware. The stated goal is a European humanoid compute supply chain for Physical AI in industrial settings.
SECO's mandate covers board design, engineering, and volume production. NEURA Robotics presents the collaboration as the industrialization of its cognitive-robot hardware within Europe, a shift from buying components to making the electronics for each machine on the continent. Because the compute architecture is decentralized, the work involves a family of computing elements distributed across the robot rather than one central computer.
The scope of the announcement matters beyond the flagship product. The compute modules are not exclusive to the 4NE1; they are meant for NEURA's broader line of cognitive robots, which means SECO's production volumes will be driven by more than one platform and by how quickly industrial customers adopt the machines.
The 4NE1 is the product drawing most attention. It is a full-size humanoid aimed at work rather than demonstration, and NEURA Robotics lists its core specifications as follows:
| Specification | 4NE1 |
|---|---|
| Height | 180 cm |
| Weight | 80 kg |
| Payload range | 10 to 100 kg |
| Top speed | 5 km/h |
The numbers describe a working machine. A payload range running from 10 to 100 kg and a top speed of 5 km/h point to material handling in production environments, where control electronics must react close to the motors and sensors that do the lifting. That operating profile is where the compute partnership becomes decisive: a robot carrying industrial weight cannot wait on a distant central processor for safety decisions.
Distributed compute sets the engineering agenda
NEURA Robotics calls its layout a "Brain + Nervous System" model. Intelligence is distributed to the joints and sensors rather than concentrated in one onboard unit, so the robot can respond in real time during physical interaction with people and equipment. The SECO-built modules carry the processing load in that distributed arrangement, and the "Smart Limbs" concept places decision-making close to the point of motion.
The rationale is latency and safety. When an arm meets unexpected resistance, the relevant controller has to react in milliseconds, and moving that decision to the limb shortens the path the data must travel. Distributed compute also makes the electronics problem harder: boards embedded in different parts of the machine must work as one system and be manufactured to consistent tolerances across every unit. Those boards run near motors and power electronics, in tight spaces, under heat and vibration, which raises the engineering bar for the manufacturer.
The choice of partner follows from that architecture. SECO brings embedded-computing experience rather than general electronics assembly, and the two companies say this is why the modules can move from prototype to industrial scale inside Europe. For buyers, the practical consequence is a robot whose electronic content has a named European manufacturer behind it from the first board to the spare parts delivered years later.
A European humanoid compute supply chain takes shape
The strategic intent is resilience. NEURA Robotics and SECO want industrial customers to source the computing core of Physical AI systems from European design and manufacturing capacity rather than depend on overseas supply lines for a component class that is becoming central to factory automation. The partnership applies to the electronics layer the same regional sourcing discipline that manufacturers already apply to robot mechanics and software. A European humanoid compute supply chain does not remove every dependency, but it concentrates the critical work in two named companies with production inside the EU.
Italy anchors the plan. A training hub called NEURA Gym is planned there to collect real-world physics data for AI training, grounding robot learning in physical experience rather than simulation alone. Real-world training data is one of the scarce inputs in humanoid development, and the facility gives NEURA Robotics a way to generate it in the same region that builds the hardware.
A second loop ties the robots back to the electronics industry. NEURA Robotics says data gathered from industrial deployments will be used to automate semiconductor and electronics manufacturing processes. Read together with the modules, that creates two returns from every installation: physics data that improves the AI, and process data that improves the manufacturing of electronic hardware. The humanoids running SECO-built compute are expected to help operate the very factories that produce electronics.
Qualcomm silicon anchors the three-way structure
The processing foundation comes from Qualcomm. Dragonwing processors sit at the heart of the SECO-built modules, and NEURA Robotics and SECO describe the agreement as an extension of a long-term strategic collaboration between NEURA Robotics and Qualcomm. In practice, the silicon relationship predates this announcement and is intended to continue across future generations of robot hardware.
Each partner holds one layer of the stack. Qualcomm supplies the processors, SECO engineers and manufactures the boards built around them, and NEURA Robotics designs the robots and the distributed architecture those boards plug into. For industrial procurement, that separation is useful: a buyer can trace the electronics to a specific manufacturer, which matters for maintenance contracts, spare part availability, and the long service life expected of factory equipment.
The humanoid market now faces a structural question that has already reshaped the wider AI industry: who controls the compute layer and who can supply it at volume. NEURA Robotics has answered with an Italian manufacturing partner and a Qualcomm processor strategy. The deal gives SECO a robotics growth path built on its embedded-computing base and keeps the manufacturing footprint inside the European Union, near the industrial customers the 4NE1 targets.
For European companies evaluating humanoid investments, this settles part of the procurement question before the robot reaches the market. The intelligence hardware will carry a regional design and production trail, with Qualcomm named as the silicon partner and SECO as the manufacturer, and industrial buyers plan around support measured in decades rather than quarters. The near-term milestones to watch are the production ramp at SECO and the build-out of NEURA Gym in Italy, both committed under this partnership.
Why this matters
Humanoid robotics is entering the phase where hardware supply chains decide which platforms reach factories at scale. Pairing Qualcomm silicon with SECO manufacturing gives NEURA Robotics a European humanoid compute supply chain that industrial customers can source regionally, while the NEURA Gym data pipeline gives the hardware a path to improve from real deployments. When the 4NE1 reaches the market, it will be judged on this infrastructure as much as on its own specifications.
Related Articles
- AWS, NEURA Partner for Cognitive Robotics Deployment
- Qualcomm Targets Hyperscale Market with CPU Built for Agentic AI Workloads
- Siemens Deploys HMND 01 Alpha Humanoid Robots with NVIDIA
✔Human Verified
Researched and cross-referenced against primary sources by the Bytevyte editorial team. This article was generated with the assistance of artificial intelligence and reviewed by the Bytevyte editorial team.