Skip to data content

Avlys OTS Data / Video data

Video training data for robotics and AI

Avlys Data helps teams specify video and observation sequences for robot manipulation, action understanding and multi-view evaluation. The catalogue describes robot demonstrations and synthetic scenes. Human egocentric capture is a separate pilot, scoped around your environment and recording requirements.

Request sample & pricingExplore collections

From Hyderabad and Bengaluru, India. Working with teams globally. Availability confirmed for your brief.

Explore the specifications

Video data collections

Compare 3 options by task, coverage and format. Request sample availability and a quote for your chosen scope.

Real robot demonstrations

Avlys OTS Data

Robot manipulation demonstrations

Multi-camera robot observations paired with actions and task context for manipulation-policy learning.

Reference collection scope76,000 trajectories · about 350 hours

564 scenes · Franka robot platform

  • Robot imitation learning
  • Visual manipulation
Dataset specifications for Robot manipulation demonstrations
Formats & labels
Video and robot-state data; release-specific training formats. Camera calibration, language instructions and robot observations / actions.
Fit & scope
Robot-camera views, not human head-mounted video. Confirm camera, action and hardware compatibility with your pipeline.
Request a quote by email: Robot manipulation demonstrations

Robot image sequences + actions

Avlys OTS Data

Tabletop robot trajectories

Robot manipulation episodes with image observations and action data across varied tabletop environments.

Reference collection scope60,096 trajectories

24 environments · WidowX robot platform

  • Language-conditioned control
  • Robot policy learning
Dataset specifications for Tabletop robot trajectories
Formats & labels
Image sequences, robot states and release-specific episode archives. Task language and robot trajectories, including teleoperated and scripted collections.
Fit & scope
Robot image sequences and actions, not ordinary stock footage or human egocentric recordings.
Request a quote by email: Tabletop robot trajectories

Synthetic multi-view video

Avlys OTS Data

Synthetic multi-view video

Physically simulated scenes filmed by virtual moving cameras, with aligned camera metadata.

Reference collection scope10,000 synthetic scenes

Up to 16 views per scene · 24 fps · 576 × 384

  • Novel-view synthesis
  • Dynamic scene understanding
Dataset specifications for Synthetic multi-view video
Formats & labels
MP4 videos; JSON metadata; NPZ camera arrays. Per-frame camera intrinsics, extrinsics and timestamps.
Fit & scope
Rendered scenes, not human-captured footage. Camera-view counts can vary by scene.
Request a quote by email: Synthetic multi-view video

Build a useful first batch

What to include
in your brief.

Start with the task your model needs to perform. A scoped pilot establishes the selection, review process and delivery expectations before you scale.

See the delivery checklist
  1. Task, environment, episode boundaries and volume
  2. Camera views, frame rate and calibration
  3. Robot platform, action format and sensor requirements
  4. Annotations, scene-separated evaluation and pipeline compatibility

Before you choose

Video data questions.

Pricing, quality & delivery FAQs
What video data options does Avlys describe?

The catalogue describes multi-camera robot demonstrations, tabletop manipulation trajectories and synthetic multi-view scenes. Check camera format, task descriptions, action representation and hardware compatibility against your pipeline. A video collection and a robot-action dataset may require different preparation.

Does Avlys offer egocentric video data for robotics?

You can discuss a first-person capture pilot with Avlys. Environment, camera placement, contributor permissions, sensors, annotations and delivery scope are agreed before collection. Robot trajectories and synthetic scenes are not human head-camera recordings, and this page does not claim a ready-to-deliver egocentric inventory.

From selection to scope

Let’s scope your
video data.

Share the task, coverage, approximate volume and timeline. We’ll discuss availability, sample arrangements and a quote for the agreed work.

Request sample & pricingDiscuss on WhatsApp ↗
sales@avlysai.com

← All Avlys Data services