Skip to content
Evora

Data

Multimodal demonstrations of skilled industrial work.

Robotics companies and labs need real demonstrations of skilled work, not generic video. Evora sources sites and workers, captures multimodal data to your spec, and delivers it in standard robotics formats — with commercial rights cleared before a camera turns on.

Request data

What you get

Real work, captured to spec.

  • Demonstrations to your spec

    You define the task and the target environment. We source the site and the workers.

  • Multimodal capture

    Synchronized RGB and depth, multi-camera angles, and hand and body pose.

  • Labels that matter

    Task and action segmentation, success and failure states, tool and object metadata.

  • Standard formats

    Delivered in standard robotics formats, ready for your training pipeline.

  • Rights cleared first

    Written consent and commercial rights from every worker and facility before recording.

  • Provenance manifest

    Consent and chain of custody documented for every delivery.

How it works

From spec to delivery.

  1. 01

    Spec the task

    Task set, environment, hours, and delivery format, agreed in writing.

  2. 02

    Source site and workers

    We match a partner facility and skilled workers, and clear consent and rights.

  3. 03

    Capture and QA

    Instrumented sessions on real equipment, reviewed against your spec.

  4. 04

    Deliver

    Synchronized, labeled data with a provenance manifest, in your format.

Deliverables

What arrives with every capture.

Each delivery ships with the same core streams and labels, plus anything your spec adds.

spec / deliverablesv1
rgb
RGB video
depth
Depth, aligned to RGB
sync
Multi-camera sync
pose
Hand and body pose
segments
Task segmentation
outcome
Success / failure states
metadata
Tool and environment metadata
provenance
Provenance manifest
format: standard robotics formats · rights: cleared before capture

Coverage

Industries and regions.

Industries
  • Warehouse and logistics
  • Industrial maintenance
  • Electronics assembly and repair
  • Data-center operations
  • Automotive service
  • Manufacturing and machine-tending
Regions
  • United States
  • East Africa
  • Mexico
  • Colombia
  • Philippines

FAQ

Questions.

What format is the data delivered in?
Synchronized RGB and depth video, multi-camera angles where applicable, hand and body pose, task and action segmentation, tool and object metadata, success/failure labeling, and a provenance manifest documenting consent and chain of custody.
Can you collect a specific task we specify?
Yes. You provide the task and target environment; we source the site and workers, capture the demonstrations, and deliver against your spec.
How do you handle consent and rights?
Every worker and every facility signs a written agreement before recording starts, covering commercial use and compensation. See our legal page for details.
What industries do you currently cover?
Warehouse and logistics, industrial maintenance, electronics assembly and repair, data-center operations, automotive service, and manufacturing and machine-tending, with more added by customer demand.
Do you work with robot policies directly, not just human demonstrations?
Yes — through our Sites program, we can run your policy in a real partner environment, identify failure modes, and generate targeted corrective data.
What's your minimum engagement?
We typically start with a scoped pilot: a defined task set, a fixed number of hours, and an agreed delivery format, before any larger commitment.

Request data

Scope a data pilot.

Tell us the task, the environment, and the volume you need. We start with a defined pilot before any larger commitment.

Inquiry: Data

We reply within 5 business days.