Egocentric Patient Room Prep and Surface Decontamination Data

It is first-person video of a person preparing a hospital patient room and wiping down its beds, rails and equipment, recorded from the operator viewpoint with depth, 3D hand pose, contact events and frame-accurate action segments on one hardware clock. The tasks are the two rows the site’s catalog lists in Hospital & Clinical Care, named "Patient room prep" and "Surface decontamination".

The data is human capture by paid, consented contributors, not robot or teleoperation data. This page describes what is collected for robot training, not how to deliver care, and nothing here is clinical guidance.

Two, both in the Hospital & Clinical Care environment under Healthcare & Life Sciences, in the catalog data file lib/commercial-catalog-data.ts. "Patient room prep" has the description "Make the bed, stock supplies, position equipment and set up the room for the next patient.", is rated Easy and lists "Phone (head-mounted) / multiview rig" as its capture setup. "Surface decontamination" has the description "Wipe down beds, rails and equipment with disinfectant and dispose of used wipes.", is rated Easy and lists the same capture setup.

Both rows are listed as Available on request. They were read on the live Hospital & Clinical Care environment page on 4 November 2026. The first table below reproduces them as the catalog states them. The site names no hospital, ward or provider for either row.

The site defines Available on request as work that is scoped and recruited to spec when a buyer asks for it. That differs from Ready to deploy, where contributors and site access are already set up, and from Active collection, where recording is running now. Neither hospital row carries either of those labels.

The site states no delivered hours, no clip count and no delivered footage for these rows, so this page claims none. What an order delivers, and when, is quoted against your spec.

The catalog states one capture setup for both rows: a head-mounted phone or a multiview rig. A head-mounted phone records the first-person view, a multiview rig adds synchronized third-person angles when the spec asks for them, and a wrist-mounted camera is available alongside the head camera for close-range detail.

Stereo is not stated on either row. It applies only through the site’s general stereo offer, where a stereo pair is a capture choice agreed in the skill spec and confirmed on a Spec Pilot. This page therefore treats stereo as an option for an order, never as the setup the catalog lists. The fourth table below sets what is stated beside what is not.

Four pages sit nearby, and none has these two rows as its main topic. The egocentric hospital and clinical care page covers all five rows in the environment, so its main topic is the whole environment. The clinical equipment setup page covers one row, Equipment setup, which positions and connects monitors, pumps and stands at the bedside.

The egocentric bed making page treats making, stripping and dressing beds as the whole subject, with the bed as the object. The stereo restroom cleaning and disinfection page covers four restroom rows in Commercial Cleaning & Janitorial. This page is the one where a patient room turnover, with its bed, rails and wipes, is the main topic. The third table below sets them side by side.

The row names four activities: make the bed, stock supplies, position equipment and set up the room for the next patient. The site states the physical steps only. It names no bed type, no supply list, no device model and no standard for what counts as a prepared room.

The four activities do not look alike on camera. Making the bed is a two-handed job on cloth, stocking is picking and placing, and positioning equipment is moving larger objects. Which of them an episode records, and in what order, is set in the skill spec.

The row names two activities: wipe down beds, rails and equipment with disinfectant, and dispose of used wipes. The site names no disinfectant, no wipe type and no order in which surfaces are wiped. It makes no claim about correct infection-control practice, and this page makes none.

In first-person video, a wipe-down is a repeated hand motion across a surface, with a hand that holds a wipe and a hand that steadies or lifts. The disposal step is a separate action. The second table below lists what each row states and what it leaves open.

A room turnover is the two rows run one after the other: the room is set up for the next patient, and the beds, rails and equipment are wiped down. The site lists them as separate rows and does not describe a combined turnover task, a required order or an episode length, so any pairing is a choice in the skill spec.

A spec can ask for each row as its own set of episodes, or for both in one sequence with action segments that mark where preparation ends and decontamination begins. That choice decides how a policy sees the boundary between the two, so settle it before capture and check it on the Spec Pilot.

The site states that 3D hand pose is delivered as two hands of 21 joints each in metric units in the head-camera frame, with a visibility flag per joint, aligned to depth and RGB. For this work, that records how the fingers hold a wipe, grip a rail or smooth a sheet, and the flag records which joints were hidden behind an object. Contact events mark when fingers make and break contact, and action segments mark where one step ends and the next begins.

The site states that egocentric video puts hands entering the frame from below, occlusion during grasp and motion blur during reach into the training distribution. A hand sliding under a rail or reaching behind equipment is that case. Force or tactile data is not stated for these rows, so treat it as an add-on to agree in the spec.

No real patients are recorded. The healthcare data page states that capture in healthcare settings happens only in training environments with simulated patients, and that real patient footage is never captured, under any circumstance, regardless of what a spec requests. The site does not say which person or object stands in for the patient on these two rows, so that is settled in the spec.

The catalog example text for this environment ends with the words "requires HIPAA compliance". That is catalog wording and nothing more. The site states no HIPAA certification, no compliance programme and no audit for these rows, and this page claims none. Because capture uses simulated patients only, the spec is where any regulatory question for your own program is raised.

For every delivered episode the site states that faces, plates and screens are detected and irreversibly blurred at ingest, that a reviewer confirms the result before anything ships, that any identifiable un-consented person goes to human review, and that footage that cannot be de-identified is discarded. Every episode links to a signed contributor release through a documented consent chain, and the consent record and the anonymization record are delivered with the data.

The site does not state how the screens category applies to a bedside monitor display, and it does not list wristbands, labels or chart text as blur categories, so this page does not claim they are blurred. Treat them as spec items and settle in the spec how the room is set up.

Each episode follows the episode schema used across Firsthand custom collections: ten streams synced under 2 ms, with head and wrist RGB, depth with distance-invalid pixels flagged, 3D hand pose with per-joint visibility, contact and grasp events, and frame-accurate action segments. Action segments carry start, end, verb and noun as the skill spec defines them.

The site states no camera count, resolution, frame rate or measurement of force or pace for these rows, and no label for a prepared room or a clean surface, so this page makes no precision claim. The fifth table below sets each stream beside what the site states about it.

Hand occlusion is the first risk. The site states that fingers disappear behind objects, and a hand closing on a rail or working under a bed frame is partly hidden from the head camera. The visibility flag on each joint is the answer the site gives for hidden fingers, and a spec can ask for wrist views so a grip the head camera loses is still recorded.

The second is that nothing in the catalog defines a finished pass. The site states no label for a prepared room or a wiped surface, so what counts as one finished pass is written into the spec. The third is the scene itself: how the room is staged, and who plays the patient, are not stated and are agreed before capture.

The catalog lines name six activities between them: make the bed, stock supplies, position equipment, set up the room, wipe down and dispose of wipes. Action segments can follow those verbs, with the surface or object as the noun. The site states no episode length, no segment count and no order, and the catalog does not split either row into finer steps.

Whether an episode is one step, one row or a full turnover is set per program. Put that boundary and the verbs you expect into the skill spec, and check them on the Spec Pilot before you scale.

Episodes are machine-checked at ingest against the stated reject criteria, and a human reviewer then checks the result. The site states these reject criteria for every episode: the manipulated object out of frame for more than 15 percent of the episode, more than 35 percent invalid depth pixels in the manipulation region, a sync breach beyond the ceiling for more than three consecutive frames, calibration drift beyond tolerance, an undocumented bystander, and a mismatch against the named skill spec.

A rejected item goes to the reject log, which ships with every batch, and is recaptured at our cost. Each batch is validated against signed accept criteria before it ships. Calibration is measured at the start and end of every session under a 0.28 px RMS gate.

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 lists clinical settings among its thinner-coverage environments, and it does not map Available on request to a specific number of days, so the timeline is quoted against your spec.

This is custom collection at scale through a network of thousands of capture providers, for a consistent, recurring data flow to labs. This page does not claim an off-the-shelf dataset of room prep or decontamination. Collections are billed per validated hour, and on-demand top-ups run against the same spec and accept criteria after each training run.

Delivery formats are RLDS, LeRobot and WebDataset, plus HDF5, zarr and Rerun .rrd. The first three are the training-loader formats, HDF5 and zarr carry arrays for custom pipelines, and Rerun .rrd gives a multi-stream synchronized playback file for review. The sixth table below sets each against what its official documentation states.

Delivery is rolling, per validated batch, and the consent record, anonymization record and reject log ship with it.

Describe which of the two rows you need, the steps and their order, the objects and surfaces involved, what counts as a finished pass, how the room starts and ends, and who or what plays the patient. We turn that into a written spec covering tasks, environment, capture tier, annotation depth, the scene boundary and accept criteria, then quote reach, timeline and price before any capture begins.

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 last table below shows what each spec element decides.