The Signal Layer for
Embodied AI
SignIQ Lab builds the data infrastructure that bridges simulation and reality. We operate a global network of instrumented facilities and expert operators to deliver production-grade robotic datasets at scale.
Our Mission
Close the data gap between simulation and the real world.
Training robots that work in unstructured environments requires massive volumes of high-quality, diverse manipulation data. Most teams can't build the capture infrastructure or hire the operators to produce it.
We solve this with a vertically integrated stack: instrumented facilities, expert teleoperation crews, automated quality pipelines, and standardized delivery formats. Researchers define what they need; we handle everything from scenario design to HDF5 packaging.
Teleoperation
Teleoperation Technology
VR Interfaces
Immersive, low-latency control for intuitive manipulation tasks.
Isomorphic Arms
Primary modality for high-precision, kinematic-accurate data.
Motion Capture
Full-body human tracking for retargeting and behavior cloning.
Sensors
Multi-Modal Sensor Suite
RGB Cameras
Head-mounted and wrist-mounted high-resolution RGB cameras.
Depth Sensors
High-fidelity depth maps for 3D spatial understanding and point clouds.
Tactile Feedback
Advanced hand-mounted sensors for force and texture perception.
Robotics
Supported Platforms
Formats
Supported Data Formats
LeRobot
RecommendedNative format for Hugging Face LeRobot framework. Optimized for immediate training.
HDF5
PopularStandard hierarchical format for high-performance storage of complex multi-modal data.
MCAP
FlexibleModern container format for multi-channel time-series data from robotics systems.
Environments
Collection Environments
Home Environment
Train robots for household assistance with realistic living rooms, kitchens, and bedrooms. Capture dual-arm manipulation for cooking, cleaning, object organization, and furniture interaction. Perfect for developing home helper robots that can autonomously cook meals, tidy spaces, and assist with daily chores.
Retail & Supermarket
Enable retail automation with fully equipped checkout counters, shelving systems, and product displays. Collect data for shelf restocking, inventory scanning, customer assistance, and product handling. Ideal for training robots to work alongside humans in grocery stores and retail environments.
Pharmacy
Develop precise medication handling with authentic pharmacy setups including medicine shelving, prescription counters, and automated dispensing systems. Train robots for error-free drug dispensing, inventory management, and pharmaceutical logistics with 99.9% accuracy requirements.
Logistics & Sorting
Master warehouse operations with industrial-scale logistics setups for bin picking, parcel sorting, and material handling. Capture high-speed sorting data (300+ items/min) and train AMRs for autonomous navigation in complex warehouse environments with dynamic obstacles.
Power & Utilities
Enable autonomous infrastructure inspection with electrical panels, substations, and utility equipment. Train quadruped and mobile robots for thermal monitoring, equipment health checks, and safety inspections in unmanned facilities, reducing downtime and improving safety.
Laboratory
Accelerate scientific research with automated lab bench tasks, pipetting, sample handling, and equipment operation. Train cobots and mobile manipulators to work alongside scientists, improving reproducibility and freeing researchers for high-value work.
Inspection
Develop autonomous monitoring systems with analog gauges, pressure meters, flow sensors, and industrial equipment. Train robots for continuous 24/7 inspection, gauge reading with computer vision, and predictive maintenance to replace episodic manual surveys.
Real-World Scenarios
Beyond our core scenarios, we continuously expand with 193 real-world environments spanning restaurants, hospitals, schools, apartments, and more. Our facility covers 376,201 sqft of diverse settings including edge cases like extreme lighting, cluttered spaces, and dynamic human interaction.