FORCE-CONTROLLED ASSEMBLY CAPTURE FOR HUMANOID ROBOTS · SOFIA, BULGARIA
The multimodal data supply for Physical AI.
Humanoid robots learn force-controlled, bimanual assembly from real human demonstrations, vision, force-torque, and tactile captured together. Roborecs records that data inside the EU in LeRobot-compatible format, the foundation for the specialist models we build next.
From human task to robot training data.
Operator Tasking
We recruit and train human operators on force-controlled, bimanual assembly, the contact-rich tasks humanoid OEMs need their robots to learn. Clients commission the specific task libraries.
TASK LIBRARY · CUSTOM COMMISSIONSMultimodal Capture
Operators perform tasks while we capture eight synchronized data channels at native rates: mocap, depth, force-torque, audio, segmentation, tracking, gaze, and localization.
8 MODALITIES · 240 Hz CAPTURE · LEROBOT v3Sync, Label & License
Sub-millisecond timestamps align every channel. Faster streams downsampled; slower ones interpolated. Output: labelled training frames at 30 Hz, ready for humanoid foundation model training pipelines.
GDPR · EU AI ACT · HDF5 / LeRobot v3This is what robotics used to be.
A programmed arm, one motion, repeated forever. Precise, but blind: change the part, the grip, or the position and it fails. Humanoid assembly is the opposite. Two hands, force-controlled, contact-rich: seating a connector, driving a fastener, placing a component with the same precision every cycle, and the judgment to adapt when nothing is ever identical twice. That generalization cannot be programmed. It is learned from thousands of labelled human demonstrations that capture the intent, the force, the micro-adjustments a machine cannot self-generate. That data is what Roborecs is built to produce.
A human moves. A robot learns.
Move your cursor, the human responds instantly. The robot mirrors with a deliberate delay, showing the latency of real-time imitation learning. Every joint angle, every force, every hesitation captured.
From Mocap to Robot Policy
Real World vs Synthetic
The team behind Roborecs.
Serial entrepreneur. Founded Internet Bulgaria 1996. Chairman, Bulgarian-Chinese Chamber of Commerce (AgiBot partnership channel).
MSc Robotics & AI, EPFL. Calibration, sensor characterization, and force-controlled manipulation for multimodal data quality.
+ Legal partner, 20-person Sofia firm covering GDPR, IP, EU AI Act compliance from day one.
Eight modalities. One labelled frame.
The multimodal supply that humanoid robots train on.
Roborecs is building the capture facility that records all eight, fuses them in real time, and ships training frames in the format the industry already uses.
Each channel captures at its native rate. Sub-millisecond timestamps align every sample. Faster channels are downsampled; slower ones interpolated. The output is a single multimodal frame, 30 times per second.
Force-controlled, bimanual assembly. Captured for humanoid training.
We specialize in contact-rich assembly, the force and two-handed coordination that gig gloves and video can't capture. Robot OEM clients commission task libraries; we capture across the categories below, with custom commissions for proprietary needs.
- →Connector mating & cable routing
- →Precision fastening with torque
- →PCB & component handling
- →Pick & place assembly
- →Quality inspection
- →Bin sorting & packing
- →Bimanual coordination
- →Tool manipulation
- →Fine alignment
- →Collaborative task handoff
- →Assistance & guiding
- →Safe proximity work
- →OEM-defined task specs
- →Bespoke capture sessions
- →Full IP assignment
Custom Commissions are available for OEM-specific task libraries. Full IP assignment. EU GDPR-compliant provenance from day one.
DISCUSS COMMISSION →
Built for physical AI. Built in Europe.
A purpose-built data capture facility at Sofia Tech Park. 1,100 m² operational footprint scaling to 5,000 m². Direct access to STEM talent, energy infrastructure, and EU logistics. Operational from Q3 2027.
At the crossroads of European deep tech and Asian manufacturing.
Sofia sits between Europe’s deep-tech labs and the Asian robotics supply chain, with the EU’s precision-manufacturing base, electronics and automotive, next door. That is where humanoids deploy first, and the natural market for EU-jurisdiction training data. The Bulgarian cost base is the operational edge; EU jurisdiction is the moat.
“We are EU jurisdiction by corporate structure, not just by data location. Under the US CLOUD Act, US-based data suppliers remain subject to US subpoena regardless of where their servers sit. EU-based humanoid customers cannot ignore this.”
EU policy is moving the same way: the June 2026 EU Technological Sovereignty Package and the proposed Cloud and AI Development Act steer public and regulated buyers toward suppliers outside third-country legal reach.
Every humanoid robot OEM needs more multimodal data than they can collect themselves.
Met June 4, 2026 in Hong Kong; partnership discussions ongoing. AgiBot has open-sourced AGIBOT WORLD 2026 and is expanding capture operations into Europe.
Academic research collaborations and pilot programs. Bulgarian-Chinese Chamber pathway to additional Chinese humanoid OEMs.
Robot OEMs and VLA labs that consume multimodal training data. EU industrials via Bosch, Schaeffler, Volkswagen, Siemens as downstream integrators.
Capture the data your robots can't collect alone.
We build EU-jurisdiction capture programs for humanoid OEMs and VLA labs, vision, proprioception, force-torque, IMU, and audio, in LeRobot-compatible format. Tell us your robot and target tasks, and we'll scope a pilot.
Start a conversation →Europe's best-funded robotics company already proved the thesis.
NEURA Robotics proved Europe can lead physical AI training. Roborecs scales the supply side, open to every humanoid maker, not vertically integrated into one hardware company.
NEURA Robotics is investing €17M into a physical AI training facility with TU Munich. A company that has raised over €1B at a ~€4B valuation dedicating this much capital to data infrastructure proves the category.
NEURA robots train on NVIDIA GR00T N1.7. Roborecs ships data in LeRobot v3 / HDF5 format, day-one compatible with GR00T, π0, and other LeRobot-based models.
NEURA Gyms capture data primarily for NEURA's own robots. Roborecs is the open supply layer, available to every humanoid OEM, not tied to any single hardware stack.
Both NEURA (Germany) and Roborecs (Bulgaria) operate under EU corporate law. EU-based humanoid customers require EU-jurisdictioned training data. US suppliers cannot provide this.
“NEURA Gyms are physical training facilities where robots practice complex tasks under controlled variability. This generates Physical AI's scarcest resource: high-quality, real-world training data.”
NEURA Robotics, April 2026
The data layer doesn't exist yet in Europe. We're building it.
Mecka AI's $60M raise on June 1, 2026 confirmed what we already knew: embodied AI data infrastructure is the next major investable category. Roborecs is building the European node.
Cumulative humanoid investment surpasses $6B. Foundation model labs start competing for physical-world training data.
Benchmark + Conviction back the household motion-capture model. Gig workers wearing sensor gloves, the US household data layer takes shape.
SoftBank, Bezos Expeditions, NVentures back Pittsburgh's universal robotics foundation model. Revenue goes $0 → $30M ARR in months.
20+ companies above $1.4B valuation. AgiBot open-sources AGIBOT WORLD 2026. The volume play is firmly Chinese.
Framework Ventures leads the round. Human-motion data capture for robot training is now a standalone fundable category, validated by institutional capital.
EU-jurisdictioned, third-party, multimodal. The open data supply layer that serves every humanoid OEM, not one hardware company.
Join us in building Europe's data supply for humanoid AI.
We capture and license force-controlled assembly data. That data is the training foundation for the specialist models we build next.
EU AI Act core obligations apply from August 2026. High-risk rules for AI embedded in regulated products, including robots, apply from August 2028 under the 2026 Digital Omnibus. Demand for EU-jurisdiction multimodal training data is accelerating faster than US-based or China-based suppliers can lawfully supply to European humanoid customers.
Get the deck and financials.
We send the full investor deck, financial model, and data-room access to qualified partners. Email us your fund and we'll respond within 24 hours.
Request the deck →