Three ways to deploy Physical AI faster.
From richly annotated datasets and simulation-ready environments to custom Alia-rig capture services.
Datasets
4 datasets — VLM & VLA
Annotation Depth: Up to 13 Modalities per Frame
Every dataset ships with a multi-modal annotation stack. VLM datasets include modalities 1–10. VLA datasets include all 13 modalities.
1
Raw Omnidirectional Exocentric Capture
Full-sphere Exocentric RGB + depth from Alia 360° sensor
360° RGB Dense Depth IMU Timestamps
ALL
2
Raw Egocentric Capture
RGB + depth from Ego camera(s)
RGB Dense Depth IMU Timestamps
ALL
3
3D Reconstruction
Dense neural 3D scene from omnidirectional input
Point Clouds Mesh Gaussian Splats 3D Layout 3D Scene Graphs
ALL
4
Spatial Semantics
Navigable paths, obstacles, zones, and surfaces
Path Maps Obstacle Class Floor Plans Zone Labels
ALL
5
Object Semantics
Per-object identity, class, pose, and attributes
3D Bounding Boxes Instance Masks SKU Labels 6-DoF Pose
ALL
6
Physics Metadata
Object mass, friction, deformability — sim-transfer-ready
Mass Estimates Friction Deformability Collision Mesh
ALL
7
Agent Tracking
People, carts, and robots — trajectories over time
Body Keypoints Trajectories Re-ID
ALL
8
Skills & Activities
What each agent is doing — pick, place, scan, stack, mop …
500+ Skills Verb–Object Pairs Temporal Spans Role Tags
ALL
9
Temporal Context
Time-of-day, traffic density, seasonal and layout variants
Rush / Off-Peak Restocking Layout Changes Lighting
ALL
10
Ego-Exo Synchronization
RGB Video Synchronization
RGB Time Sync Ego-Exo Action Mapping
ALL
11
Hand Pose & Trajectory
Frame-level manipulation actions from egocentric view
Gripper State Contact Events Hand Pose Force Proxies
VLA ONLY
12
Body Pose & Trajectory
Frame-level manipulation actions from egocentric view
Gripper State Contact Events Hand Pose Force Proxies
VLA ONLY
13
Human to Robot Retargeting
Conversion of human trajectories to standard robot joint data
Robot Motion Robot Control Supports Unitree G1, Fourier and many more
VLA ONLY
Modalities 1–10 — included in all datasets (VLM & VLA)
Modalities 11–13 — VLA datasets only (action data)
VLM: Grocery Frontend Operations 500 hrs · 10 modalities

Customer-facing grocery store operations captured with dual camera system — Alia 360° sensor and Egocentric camera. Covers shopping aisles, checkout lanes, deli counters, produce sections, and self-checkout areas across multiple stores and time-of-day conditions.

10
annotation modalities per frame
Modalities 1–10 from the stack above
Duration
500 hours
Stores
5 operational
Camera
Alia 360° Exocentric + Egocentric
RGB + dense depth
Skills Captured
250+
Shopping, checkout, browsing
Zones
Aisles, checkout, deli, produce
Formats
Open X-Embodiment
Navigable Path Maps Obstacle Classification Customer Trajectories Temporal Variants 3D Scene Reconstruction Instance Segmentation
VLM: Grocery Backend Operations 500 hrs · 10 modalities

Back-of-store operations — receiving docks, cold storage, stockrooms, and restocking workflows. Captures the logistics side of grocery with forklift activity, pallet handling, and inventory management tasks.

10
annotation modalities per frame
Modalities 1–10 from the stack above
Duration
500 hours
Stores
5 operational
Camera
Alia 360° Exocentric + Egocentric
RGB + dense depth
Skills Captured
250+
Stocking, receiving, cold chain
Zones
Dock, stockroom, cold storage
Formats
Open X-Embodiment
Pallet Tracking Forklift Trajectories Inventory Workflows Cold Chain Operations Shift Transitions Loading Dock Activity
VLA: Grocery Frontend Operations 500 hrs · 13 modalities · dual-stream

Action-centric dataset combining Alia 360° overhead capture with GoPro egocentric cameras on workers. Includes all 10 VLM modalities plus manually corrected hand pose, body pose and human to robot retargeted data — purpose-built for training Vision-Language-Action models.

13
annotation modalities per frame
All 13 modalities including action data
Duration
500 hours
Stores
5 operational
Cameras
Alia 360° + GoPro
Dual-stream synchronized
Skills Captured
300+
With action primitives
Zones
Aisles, checkout, deli, produce
Formats
LeRobot, Open X-Embodiment
Egocentric Video Hand Pose Tracking Gripper State Labels Contact Events Force Proxies Physics Metadata Sim-Transfer Ready
VLA: Grocery Backend Operations 500 hrs · 13 modalities · dual-stream

Backend logistics with full action annotation — receiving, stocking shelves, operating pallet jacks, and managing cold-chain inventory. Dual-stream capture enables training VLA models on the physical manipulation tasks behind grocery operations.

13
annotation modalities per frame
All 13 modalities including action data
Duration
500 hours
Stores
5 operational
Cameras
Alia 360° + GoPro
Dual-stream synchronized
Skills Captured
300+
With action primitives
Zones
Dock, stockroom, cold storage
Formats
LeRobot, Open X-Embodiment
Egocentric Video Heavy-Object Manipulation Pallet Jack Actions Contact Events Mass & Friction Estimates Physics Metadata Deformability Labels
Simulation Assets
2 asset types — USD format

Simulation-ready grocery environments and objects in Universal Scene Description (USD) format, built from DreamVu's 3D reconstructions. Drop directly into NVIDIA Isaac Sim or Omniverse for robot training, validation, and sim-to-real transfer.

1
Complete USD Store Environments
5 photorealistic digital twin grocery stores — full aisle layouts, checkout areas, stockrooms, and deli counters
2
2,000+ Individual Product Assets
Grocery items across produce, dairy, packaged goods, and frozen — each with accurate geometry and PBR textures
3
Physics Properties per Object
Mass, friction coefficients, deformability, and collision meshes — ready for contact-rich manipulation simulation
4
Shelf Planograms & Aisle Topology
Real product placement layouts with navigable aisle graphs — matches actual store configurations
5
Lighting & Layout Variants
Multiple lighting presets (daylight, fluorescent, night) and seasonal layout configurations per store
6
SKU & Barcode Metadata
Product identity, category labels, and barcode data attached to every asset for scan-and-pick training
7
Grasp Point Annotations
Pre-computed grasping points and approach vectors for each product asset — accelerates manipulation policy training
8
Isaac Sim & Omniverse Compatible
Native USD/USDA format — load directly into NVIDIA simulation tools with zero conversion overhead
Custom Capture
On-site data collection & processing

DreamVu deploys to your facility with our Alia 360° capture rigs and full annotation pipeline. You get the same multi-layer annotation stack applied to your specific environment, operations, and use cases.

📷
On-Site Capture
Alia 360° rigs deployed to your location — warehouses, factories, retail, hospitals, or any operational environment.
🎯
Custom Annotation
Full 7- or 9-layer annotation stack tailored to your domain-specific skills, objects, and workflows.
🛠
3D Reconstruction
NuRec pipeline produces dense 3D scenes — point clouds, meshes, and Gaussian splats from your facility.
USD Digital Twin
Your environment converted to simulation-ready USD assets for Isaac Sim and Omniverse workflows.
📊
Model Fine-Tuning
Optional foundational model fine-tuning on your custom dataset for domain-specific world models or robot policies.
🔒
Exclusive License
Custom capture data is exclusively yours — never shared, sublicensed, or added to our catalog without your consent.

Ready to build with real-world data?

Talk to our team about datasets, simulation assets, or custom capture for your environment.

Talk to Us