Egocentric AV Cable Management and Taping Data

It is first-person video of a person managing cables at an event venue, recorded from the worker’s own point of view with synchronized sensor streams alongside. The two catalog rows that name this work describe running cables to AV gear and taping them down, and laying cables along walls and walkways, covering them with ramps and taping them down with gaffer tape.

The data is human capture by paid, consented contributors, not robot or teleoperation data, so using it for a specific robot still needs your own retargeting step. The site states one task instruction per episode as its language data. Firsthand is the collector; this page does not name any venue, event or AV equipment maker.

Two, listed in lib/commercial-catalog-data.ts, both in the Event Venue environment of the Hospitality & Events sector. "AV & Cable Management" is rated Medium and described as "Run cables to AV gear, tape them down, position speakers and screens and connect equipment." "Route and tape AV cables using gaffer tape and cable ramps" is rated Easy and described as "Lay cables along walls and walkways, cover them with ramps and tape them down with gaffer tape."

Both are listed as Ready to deploy, with "Phone (head-mounted) / multiview rig" as capture setup. The site states no delivered hours for them, so this page claims none. The environment examples read "Conference halls, wedding venues, stages and banquet rooms", which are kinds of room, not named venues. The first table below reproduces the rows as the catalog states them.

This page is the dedicated task page for the two AV cable rows. The egocentric event venue setup and teardown page covers all eight Event Venue rows, tables and chairs, staging, stage and equipment, cables and sweeps, and gives the two cable rows one short section. The egocentric cable routing and wiring page covers fifteen cable and wiring rows across four sectors, bench, rack, venue, construction and trade rows, and lists these two rows in its table.

The stereo layout marking and taping page is about measuring, marking, drywall taping and masking, not cable taping. No page other than this one is about AV cables taped under ramps. The second table below sets the pages side by side, with each status as the site lists it.

A head-mounted phone records the first-person view of real work on site, and a multiview rig adds synchronized third-person angles when the spec asks for them. Both rows list this single setup. On the rig tier, every stream is hardware-triggered onto one PTP clock domain, and the rig is calibrated before capture starts and after it ends. An episode is accepted only if RMS reprojection stays under 0.28 px.

Camera count, resolution and frame rate for venue work are Not stated, so session length and volume are quoted against your spec. The site does not state a head-mounted stereo setup for these rows; stereo is available only through the site general stereo offer, agreed in the skill spec.

The rows give a natural cut. For the first row: run cables to AV gear, tape them down, position speakers and screens, connect equipment. For the second row: lay cables along walls and walkways, cover them with ramps, tape them down with gaffer tape. The stages are the catalog’s own wording.

A spec that says "manage the cables" gives one long action segment. A spec that names the stages gives a sequence you can slice by step. The site does not set an episode length or say how many runs make up an episode. Contact events can mark when a hand meets a cable, a ramp or the tape where the spec asks for them.

The first row is broader and rated Medium: it ends with positioning speakers and screens and connecting equipment, so it includes work that is not cable laying. The second row is narrower and rated Easy: it is the floor-level run, ramps and gaffer tape. Whether one episode covers one run or a whole room is set in the spec.

Which gear is connected, what a connection involves, cable types, lengths and routes, ramp and tape types, and what counts as a finished run are Not stated. The site lists no success label for these rows.

Each episode follows the Firsthand episode schema: ten streams synced under 2 ms, with head and wrist RGB, depth, 3D hand pose with per-joint visibility, contact and grasp events, and frame-accurate action segments. Object labels use the terms your spec sets, and one task instruction is attached per episode.

Force or tactile data, and success or quality labels, are Not stated for these rows. Tactile gloves are a separate capture tier with their own page. The third table below sets each stream beside what the site states.

Hand and tool occlusion is the first risk. The site states that fingers disappear behind objects, which is why each hand joint carries a visibility flag, and that reach and grasp frames are where blur and occlusion show up. The head camera is global-shutter, which avoids rolling-shutter skew. How often a finger is hidden on cable work is Not stated.

An episode is rejected when the manipulated object is out of frame for more than 15 percent of it, when more than 35 percent of depth pixels in the manipulation region are invalid, for a sync breach beyond the 2.0 ms ceiling for more than three consecutive frames, for an undocumented bystander, or for a mismatch against the named skill spec. A reject log ships with every batch. The fourth table below lists the failure modes beside what the site states.

Protective equipment, lifting, trip-hazard rules, electrical safety, AV rigging and load details, and attendance and capacity rules are Not stated, and this page makes no safety or compliance claim. Cable types, lengths, connection standards, signal specifications, AV gear, ramp and tape brands, venue names, task times and cycle counts are Not stated too.

For every delivered episode the site states that faces, plates and screens are blurred irreversibly, that a reviewer confirms the result before anything ships, and that footage that cannot be de-identified is discarded. The anonymization record ships with the data, and every episode links to a signed contributor release. The site states no venue-specific bystander rule, so settle it in the spec before capture. This is not legal advice, and the executed agreement controls.

The October 2026 off-the-shelf catalog lists ~324,800 hours across 18 datasets. The site names no event venue, AV cable or cable ramp dataset among them; its only event venue mention in that area is 360° virtual tour imagery of spaces, not task video. This work is therefore custom collection, not an off-the-shelf purchase.

Where coverage is deep, the median time from signed spec to first validated batch is a median of 48 hours; where it is thin, turnaround runs three to four weeks. The site does not say which applies to venue work, so the timeline is quoted against your spec. Both rows are Ready to deploy, so contributors and site access are set up and capture starts once a spec is signed. Collections are billed per validated hour, and rejected episodes and recaptures are never billed.

Describe the stages and where each starts and ends: running cables to gear, laying cables along walls and walkways, covering with ramps, taping with gaffer tape, positioning speakers and screens, connecting equipment. Say whether an episode covers one stage or the whole job, which of the two rows you want, and what counts as the object and a completed step. Then name the streams you want and the delivery format: RLDS, LeRobot, WebDataset, HDF5, zarr or Rerun .rrd.

Most programs begin with a Spec Pilot, a small first batch that proves the spec and accept criteria on real footage before you scale. The fifth table below shows what each spec element decides.