Request sample access
This repository is publicly accessible, but you have to accept the conditions to access its files and content.
This listing is a preview. The production dataset (50,000 hours of first-person industrial and warehouse video with structured annotations) is rights-cleared and delivered directly under a commercial license. Approved requesters get the full annotation schema and data dictionary in this repository, and can request a review package with real video samples and QA summaries. Requests are reviewed within 1 business day.
Log in or Sign Up to review the conditions and access this dataset content.
Industrial First-Person Video Dataset
50,000 hours of egocentric workplace video covering picking, packing, scanning, stocking, inventory, sorting, inspection, returns, assembly, and cleaning workflows for Physical AI.
This repository is a specification and preview listing. The production dataset is rights-cleared and delivered directly to buyers. Request access to see the full schema and get real video samples.
Overview
The Industrial First-Person Video Dataset is a 50,000-hour collection of egocentric workplace video capturing real industrial, warehouse, logistics, retail, manufacturing, and commercial operations tasks from the human point of view. It is designed for Physical AI, robotics, industrial AI, workplace automation, safety monitoring, operations intelligence, and egocentric action recognition.
Because the footage is recorded first-person, it preserves what workplace automation models actually need: hand-object interactions, tool usage, spatial navigation, workflow order, object placement, and real-world task constraints. The dataset captures how workers move through operational environments, interact with tools and equipment, scan barcodes, pick and pack items, handle boxes, stock shelves, inspect products, sort packages, process returns, perform inventory checks, complete assembly steps, and clean commercial spaces.
At a glance
| Total volume | 50,000 hours |
| Perspective | Egocentric (first-person) |
| Video quality | 1080p+ where available |
| Environments | Warehouses, logistics facilities, retail backrooms, manufacturing floors, commercial cleaning sites, quality inspection areas |
| Key events | Scan events, pick/place events, workflow phases, completion outcomes |
Task coverage
Warehouse picking and packing, barcode scanning, shelf stocking, pallet handling, inventory checks, forklift-adjacent workflows, package sorting, quality inspection, returns processing, manufacturing assembly, janitorial workflows, commercial cleaning, equipment checks, and workplace organization.
Technical specifications
- Annotation coverage: task type, action segments, object labels, tool labels, location zones, hand-object interactions, scan events, pick/place events, workflow phases, completion outcomes, QA review, privacy redaction
- Buyer evaluation metrics: task diversity, location diversity, hand visibility, object visibility, scan-event coverage, workflow completion labeling, safety-event labels where available
- QA metrics: blur/exposure score, usable video percentage, annotation quality score, privacy redaction review, duplicate removal, incomplete workflow filtering
- Delivery format: MP4 video, CSV metadata, JSON annotations, thumbnails or preview frames
Annotation schema
The gated file annotation_schema.json in this repository contains the full clip schema with an illustrative example record, including workflow phases, scan and pick/place events, and privacy-redaction QA fields.
How to evaluate this dataset
- Request access using the form above. Requests are reviewed within 1 business day.
- On approval you get the gated files in this repository: full annotation schema, data dictionary, and access instructions.
- Request a review package and we deliver real video samples in your target environments, JSON annotations, metadata CSVs, QA summaries, and licensing documentation within 2 business days.
All samples are delivered with structured CSV metadata and JSON annotation files where available. Buyer review packages include representative media files, metadata samples, annotation schema, QA summaries, and data dictionary documentation.
Licensing
The production dataset is rights-cleared for commercial AI training and licensed directly by Datoric, with documented worker and site consent, privacy redaction review, and chain-of-custody. Subset (by environment, task category, or hours), exclusive, and custom-collection options are available.
About Datoric
Datoric supplies rights-cleared, spec-exact training data for frontier AI labs and enterprise model teams: industrial and egocentric video, human manipulation data, robot episodes, expressive multilingual voice, computer-use traces, and gameplay trajectories. We also run managed collection pipelines for custom specifications.
Contact: nikhil@arzule.com
- Downloads last month
- 12