Now recruiting capture teams worldwide. Get paid to record first-person video of real work.

Apply to capture

Hard task

Egocentric Clinical Procedures on Mannequins Data

Dressings, injections and checks performed on a phantom or mannequin: the task catalog lists them as one row, rates it Hard, and this page covers that row on its own. It explains what egocentric clinical procedure data is, how the row is listed and captured, how it differs from the neighbouring first aid, hospital and sterile field pages, and what the site does not state. We run custom clinical data collection fast and on demand against a spec you write, as first-person phantom and mannequin training video and a dressing and injection practice robot dataset, as custom collection at scale through a network of thousands of capture providers, for a consistent, recurring data flow to labs. It is practice on a phantom or mannequin, never a real patient.
Catalog row
"Medical Procedures" in Hospital & Clinical Care, rated Hard
Status
Available on request, as the catalog lists it
Capture setup
Phone (head-mounted) / multiview rig, as the catalog lists
Setting
Phantom or mannequin practice only; no real patients
License
Perpetual, buyer-owned

Requirements

What does clinical procedure practice data need?

01

One row, three kinds of activity

The row names dressings, injections and checks on a phantom or mannequin. A spec says which of them a collection records, rather than relying on one clinical label.

02

Both hands and the objects in frame

A head-mounted view keeps the hands and the objects they hold in view through each step, with a wrist camera available alongside for close-range detail.

03

Hand pose and step boundaries on one clock

3D hand pose with a visibility flag per joint and frame-accurate action segments are aligned to video and depth on one hardware clock, ten streams synced under 2 ms.

04

A scene boundary settled before capture

The site states a simulated-patient-only boundary for healthcare settings and never captures real patient footage. Which phantom or mannequin stands in for the patient is agreed in the written spec.

Definition

What is egocentric clinical procedures on mannequins data?

It is first-person video of a person carrying out clinical procedures on a phantom or mannequin, recorded from the operator viewpoint, with depth, 3D hand pose, contact events and frame-accurate action segments aligned on one hardware clock. People reach it through several phrasings: egocentric clinical procedure data, first-person phantom and mannequin training video, a dressing and injection practice robot dataset, or custom clinical data collection at scale.

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 perform a procedure. It gives no instructions, makes no clinical-accuracy, safety or outcome claim, and states no result for any robot. Using the data for a specific robot still needs your own retargeting step.

Catalog row

Which catalog row covers clinical procedures on a mannequin, and what is its status?

One. "Medical Procedures" sits in the Hospital & Clinical Care environment, in the Healthcare & Life Sciences sector. Its description reads "Perform clinical procedures on a phantom or mannequin, dressings, injections and checks." It is rated Hard, it lists "Phone (head-mounted) / multiview rig" as the capture setup, and it is Available on request. The first table below reproduces it exactly.

Available on request means the work is scoped and recruited to spec when a buyer asks for it, rather than running as an open capture stream today. The site states no delivered hours for this row, so this page claims none. The row names a phantom or mannequin and three kinds of activity, and it names no phantom type, no mannequin type, no dressing, no injection route and no check, so anything more specific is written into the spec and not assumed.

Difference

How is this different from the hospital, first aid and sterile field pages?

The egocentric hospital and clinical care page is the family page: it covers five rows, of which Medical Procedures is one, so this page isolates that row. The egocentric first aid and clinical skills page covers five different rows, First Aid Training, BP cuff (mannequin), Wound dressing (phantom), Sterile field setup and Instrument pass, and it states that Medical Procedures sits close to its five but is not counted. The sterile field setup and wound dressing page covers three Wet Lab & Clinical Bench rows, all rated Medium and all listed with a multiview rig plus handheld grippers.

Medical Procedures is a separate row in a separate environment, rated Hard, with a head-mounted phone or multiview rig as its capture setup. This is the one page where that row is the whole subject. The second table below sets the pages side by side.

Hard

What does Hard mean for this row, and why is it demanding?

Hard is the skill level shown in the catalog, one of Easy, Medium or Hard. The site states no further criteria, so this page adds none. The only grounds for calling the row demanding are in its own wording: it bundles dressings, injections and checks into one row, it is performed on a phantom or mannequin, and the catalog rates it Hard while rating two of the other four hospital rows Easy and two Medium.

The site states no failure rate, tolerance, cycle time or skill level for the people who perform it, and this page makes no such claim. It also does not say that the row is harder to capture than another Hard row. What it does state is the mechanism that applies to any two-handed task in this family: 3D hand pose carries a per-joint visibility flag, so a finger hidden behind an object is marked hidden rather than filled in, and action segments mark where one step ends and the next begins.

Capture setup

How is the row captured?

The catalog lists "Phone (head-mounted) / multiview rig" for this row. A head-mounted phone records the first-person view, and a multiview rig adds synchronized third-person angles when the spec asks for them. A wrist-mounted camera is available alongside the head camera for close-range detail. The catalog does not say which of its two modes applies to a given order, so that is agreed in the spec.

The site describes this capture mode in general as the first-person view of real work on site, as opposed to the staged handheld-gripper mode. For this row the practice is on a phantom or mannequin, so this page does not describe it as work with real patients. The site does not state a camera count, resolution, frame rate or any measurement of force or pace for this row, so this page makes no precision claim. The third table below lists the setup items.

Who is in the scene

Are real patients recorded?

No. 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 row itself says phantom or mannequin. No patient data and no clinical outcome is part of the collection.

The healthcare data page also describes its task set as deliberately non-clinical and says the site does not perform or claim to perform later-stage clinical validation. Medical Procedures is a catalog row for practice on a phantom or mannequin, and the row wording is the only description the site gives. The catalog example text for the 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 this row, and this page claims none.

Episode contents

What does each episode contain?

Each episode follows the episode schema used across Firsthand custom collections: 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. Depth flags distance-invalid pixels rather than interpolating them.

For a procedure on a mannequin, contact events mark when the fingers make and break contact with an object, and action segments mark where one step ends and the next begins. Whether a dressing, an injection or a check is its own segment, and whether a whole procedure or one step is one episode, is a spec decision. The site does not set an episode length or a step count. The fourth table below sets each item against what the site states.

Not stated

What does the site not state about this row?

Which steps are recorded, which phantoms or mannequins are used, any named dressing, injection route or check, the tools and materials beyond the words in the row, episode lengths, error rates, force or tactile readings, any success or completion label, any sterility standard, any licensing or regulatory status, and what a correct procedure means are Not stated. They are set in the written spec, and this page claims none of them.

The site also names no hospital, clinician, facility or client for this row. Who performs the task and where it is recorded are agreed in the spec, and capture at a customer’s own facility is listed as not offered.

Privacy

How are consent and privacy handled?

Every episode links to a signed contributor release through a documented consent chain, and the consent record is part of the delivery. Faces, plates and screens are detected and irreversibly blurred at ingest, and a reviewer confirms the result before anything ships. Clips that cannot be de-identified are discarded, and the anonymization record is delivered with the data.

The site does not list wristbands, labels or monitor displays as blur categories, so this page does not claim they are blurred. Treat any such item as a spec item and settle in the spec how the scene is set up.

Formats

Which delivery formats carry this data?

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 fifth table below sets each against what its official documentation states.

Every batch is validated against signed accept criteria before it ships, with a reject log, and delivery is rolling, per validated batch.

Fast and on demand

How fast can a custom clinical procedure dataset be collected, and at what scale?

This is custom collection at scale through a network of thousands of capture providers, for a consistent, recurring data flow to labs. 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, and the site lists clinical settings among the thin environments.

The row is Available on request, and the site does not map that status to a specific number of days or state a contributor count for it, so the timeline is quoted against your spec. On-demand top-ups run against the same spec and accept criteria after each training run. Collections are billed per validated hour, and rejected episodes and recaptures are never billed.

Ordering

How do you order a clinical procedure dataset on a mannequin?

Describe which of the three activities the row names you need, where each episode starts and ends, what stands in for the patient, the setting and lighting, and the privacy rule for anyone in frame. Name the variation your pipeline is missing, whether a wrist view or a multiview angle is needed, and which delivery format your loader expects.

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

Robots

Why does this data matter for robot learning?

Practice procedures on a phantom or mannequin are sequences of two-handed manipulations in a set order. First-person data gives the hand and object detail for each move, and the action segments give the order. That is the useful part for a dressing and injection practice robot dataset.

The site states no robot results for this row, makes no claim that any model was trained on this data, and makes no safety claim about robots working near patients. This page describes what is collected, not what a policy trained on it will do.

At a glance

What you get.

Catalog row
"Medical Procedures" (Hospital & Clinical Care, Healthcare & Life Sciences); Hard; Available on request
Catalog description
"Perform clinical procedures on a phantom or mannequin, dressings, injections and checks."
Catalog capture setup
Phone (head-mounted) / multiview rig
Environment examples line
"Patient room prep, equipment setup, specimen handling, surface decontamination -- requires HIPAA compliance"; catalog wording only, no certification is claimed
Episode
Ten streams synced under 2 ms; head and wrist RGB; depth with invalid pixels flagged; 3D hand pose with a per-joint visibility flag; contact and grasp events; frame-accurate action segments
Setting boundary
Training environments with simulated patients only; real patient footage is never captured, regardless of what a spec requests
Delivery formats
RLDS, LeRobot, WebDataset, HDF5, zarr and Rerun .rrd
Billing and QA
Per validated hour; signed accept criteria; reject log per batch; rejected items recaptured at the provider’s cost
Which steps, which phantom or mannequin, durations, tools, outcomes
Not stated; agreed in the skill spec
License
Perpetual, irrevocable, buyer-owned, non-exclusive by default

Every collection runs through the same seven-stage end-to-end custom collection pipeline. End-to-end custom data collection is a managed service that takes an AI data need from problem to owned dataset in a single accountable pipeline.

Catalog row

How does the catalog state the Medical Procedures row?

One row from the task catalog, reproduced as the catalog states it, plus items the catalog does not list.
Task (as listed)Description (as listed)EnvironmentSkillCapture setupStatus
Medical ProceduresPerform clinical procedures on a phantom or mannequin, dressings, injections and checks.Hospital & Clinical CareHardPhone (head-mounted) / multiview rigAvailable on request
Separate dressing, injection or check rows in this environmentNot stated as separate rowsNot statedNot statedNot statedNot stated
Named phantoms, mannequins, dressings, tools or proceduresNot statedNot statedNot statedNot statedNot stated

Compared pages

How does this page differ from the neighbouring clinical pages?

Each page and what it covers. Statuses are as the site lists them; none of these pages is claimed to hold delivered hours here.
PageWhat it coversRow status as listedRelation to this page
This page: clinical procedures on mannequinsOne row: Medical Procedures (Hospital & Clinical Care)Hard; Available on requestThe whole subject
Egocentric hospital and clinical care task dataFive rows: Equipment setup, Medical Procedures, Patient room prep, Specimen handling, Surface decontaminationTwo Easy, two Medium, Medical Procedures Hard; all Available on requestThe family page; Medical Procedures is one of five
Egocentric first aid and clinical skills training dataFive rows: First Aid Training, BP cuff (mannequin), Wound dressing (phantom), Sterile field setup, Instrument passEasy and Medium; all Available on requestLists Medical Procedures as a nearby row it does not count
Egocentric sterile field setup and wound dressing dataThree Wet Lab & Clinical Bench rows: Sterile field setup, Wound dressing (phantom), Connect Luer syringeAll Medium; Available on request; multiview rig + handheld grippersDifferent environment, rating and capture setup
Egocentric Luer syringe connection dataOne row: Connect Luer syringe, "Attach and lock a Luer syringe to a line."Medium; Available on requestA separate bench row, not this one

Capture setup

What does the capture setup for this row look like?

Left to right: the setup item, what the site states, and what that means for an order. "Not stated" means the site does not say.
ItemWhat the site statesFor a clinical procedures order
Catalog capture setupPhone (head-mounted) / multiview rig on the rowWhich mode applies is agreed in the spec
Head-mounted phoneRecords the first-person viewKeeps the hands and held objects in view
Multiview rigAdds synchronized third-person angles when the spec asksOptional angles, named in the spec
Wrist-mounted cameraAvailable alongside the head camera for close-range detailNamed in the spec if needed
SceneTraining environments with simulated patients onlyA phantom or mannequin; never a real patient
Camera count, resolution, frame rateNot stated for this rowAgreed in the spec
Force, pace or duration measurementNot statedNot claimed on this page

Streams and labels

What does each episode carry, and what is not stated?

Streams and labels the site states for Firsthand custom collections, and fields for this row that the site does not state.
ItemStatus on this siteWhy it matters for this row
Ten streams synced under 2 msStatedA step boundary lines up with the exact video and depth frames
Depth, with distance-invalid pixels flaggedStatedUnreliable pixels are marked, not interpolated
3D hand pose with per-joint visibilityStatedA finger hidden behind an object is marked hidden
Contact and grasp eventsStatedMarks when the hands make and break contact with an object
Frame-accurate action segmentsStatedMarks where one step ends and the next begins
Which phantom or mannequin, and which stepsNot statedSettled in the spec before capture
Success, completion or accuracy labelsNot statedNo claim is made
Clinical outcomes, certification, safety or robot resultsNot statedNo claim is made

Formats

Which formats carry this data?

Delivery formats stated on the site, set against what each format’s official documentation states where the documentation was read.
FormatWhat the official documentation statesUsed for here
RLDSA dataset is retrieved as episodes, and each episode holds a dataset of steps (RLDS repository README)Episode-based training loaders
LeRobotLeRobotDataset v3.0 is a standardized format for robot learning data, including multi-camera video, with many episodes stored per Parquet or MP4 file (LeRobot documentation)LeRobot dataset format and tooling
WebDatasetTar files in which files that belong together share a basename, split into numbered shards (WebDataset README)Sharded streaming for large training runs
HDF5 and zarrStated by Firsthand onlyArray storage for custom pipelines
Rerun .rrdStated by Firsthand onlyMulti-stream synchronized playback for review

Skill spec

What does the skill spec fix for this order?

Spec element and where a buyer can see it in the delivery. "Set per program" means the site states no fixed value.
Spec elementWhat it decidesWhere it shows up
RowMedical Procedures, worded as the catalog words itSet per program; status as listed
ActivitiesWhich of dressings, injections and checks are recorded, and in what orderAction segments; set per program
Episode scopeOne step, one procedure or a run of proceduresSet per program; the site does not state an episode length
Stand-in for the patientWhich phantom or mannequin is usedSet per program; never a real patient
Rig and streamsHead-mounted phone, multiview angles, wrist views, hand pose and any othersThe streams you can read; confirm on the Spec Pilot
Delivery formatHow the data is packagedRLDS, LeRobot, WebDataset, HDF5, zarr or Rerun .rrd
Accept criteriaWhat counts as a recorded step and what passes QAReject log per batch
PrivacyHow anyone, faces, screens and labels in frame are handledFaces, plates and screens blurred at ingest; other items agreed in the spec
LicenseWho owns the datasetPerpetual, irrevocable, buyer-owned, non-exclusive by default

Example footage

What does first-person capture look like?

One real clip already published on this site, shown as a format example: ten seconds of first-person home cooking with faces blurred. It does not show a clinical procedure, a phantom or a mannequin.

egocentric_cooking_10s.mp4 · first-person view · faces blurred · 1080 × 1080 web preview · format example, not a clinical procedure clip

No clinical procedure clip is published on the site, so this is labelled as a format example. Request a sample to see delivered files. See sample data outputs, request an egocentric data sample.

FAQ

Egocentric clinical procedure data, answered.

What is egocentric clinical procedures on mannequins data?
First-person video of a person performing clinical procedures on a phantom or mannequin, dressings, injections and checks, with depth, 3D hand pose and action segments, collected to a written spec. Firsthand handles contributors, capture, QA, consent and delivery.
Which catalog row covers it, and what is its status?
One, in Hospital & Clinical Care: "Medical Procedures" (perform clinical procedures on a phantom or mannequin, dressings, injections and checks). It is rated Hard and is Available on request. This page claims no delivered hours.
Does this include real patients?
No. The site states that healthcare capture uses training environments with simulated patients only and never captures real patient footage. The row names a phantom or mannequin. No patient data or clinical outcome is part of it.
How is this different from the first aid and sterile field pages?
The first aid page covers five other rows and lists Medical Procedures as a nearby row it does not count. The sterile field page covers three Medium bench rows with a multiview rig plus handheld grippers. This page covers the one Hard Medical Procedures row.
How is the row captured?
The catalog lists 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, and which mode applies is agreed in the spec.
Does the site claim clinical accuracy, safety or HIPAA compliance?
No. The catalog example text for the environment says "requires HIPAA compliance", and that is wording only. The site states no certification, no clinical-accuracy or safety result and no robot result for this row, and nothing here is clinical guidance.
How fast can a custom dataset be collected?
Where coverage is deep, the first validated batch lands in a median of 48 hours from signed spec. Thin environments typically take three to four weeks, and the site lists clinical settings among them. The row is Available on request, so the timeline is quoted against your spec.
What formats and license come with it?
RLDS, LeRobot, WebDataset, HDF5, zarr and Rerun .rrd. The license is perpetual, irrevocable and buyer-owned, non-exclusive by default, and billing is per validated hour. Faces, plates and screens are irreversibly blurred at ingest.

Sources

Where are the delivery formats documented?

Request a custom clinical procedure dataset.

Tell us which of dressings, injections and checks you need, where each episode starts and ends, what phantom or mannequin stands in for the patient, the streams you want and the format your loader expects. We will quote reach, timeline and price against a written spec, with a first validated batch proven on a Spec Pilot.