Real-world capture programs

These are example capture programs, not products in a catalog. Each one is a custom robotics data collection engagement scoped to a client's robot, sensors, environments, and evaluation goals, run from pilot to production with documented provenance and data ownership. Use them as starting points to scope your own program.

• Scenario

  AMR aisle navigation with dynamic obstacles

  Fleet-relevant AMR runs with human and equipment interference in warehouse aisles, captured with intervention and replan metadata.

• Scenario

  Autonomous driving rain and low-light LiDAR-camera

  On-road or closed-course AV capture emphasizing weather and lighting tails, with intervention and disengagement metadata.

• Scenario

  Bin Picking Teleoperation Capture

  Custom teleoperated bin picking capture for warehouse robots handling cluttered totes, occluded SKUs, and grasp recovery.

• Scenario

  Defective SKU pick failure library

  Structured failure sets for odd-shaped or damaged packaging picks, with a labeled failure taxonomy for evaluation.

• Scenario

  Dual-arm kitchen manipulation in homes

  Bimanual kitchen task capture in residential layouts with contact-rich interaction, calibrated multi-view video, and object labels.

• Scenario

  Humanoid home navigation and reach

  Whole-body navigation and reach capture in furnished home layouts, including flooring transitions, narrow passages, and fall/recovery episodes.

• Scenario

  Humanoid stair ascent and descent

  Stair locomotion capture with balance-critical recovery data across varied tread dimensions and handrail availability.

• Scenario

  Industrial peg-in-hole with force traces

  Contact-rich insertion capture with force-torque and vision-occlusion labels, including alignment errors and jam events.

• Scenario

  Laboratory pipetting dexterity capture

  Precision manipulation capture for lab automation prototypes, with sub-millimeter trajectories and contact/alignment failure labels.

• Scenario

  Outdoor rough-terrain locomotion

  Legged locomotion capture on uneven outdoor surfaces with synchronized IMU, vision, and LiDAR, including slip and recovery behaviors.

• Scenario

  Retail shelf restocking teleoperation

  Shelf-facing pick-and-place capture with SKU diversity in retail aisles, including occlusion and crowded-shelf episodes.

• Scenario

  Warehouse conveyor handoff episodes

  Robot-to-conveyor transfer capture in distribution centers with timing-critical grasp and line-speed metadata.

• Scenario

  Warehouse pallet pick teleoperation capture

  Teleoperated pallet picking with multi-camera RGB-D in live warehouse aisles, captured with grasp-outcome and SKU metadata.