Egocentric Childcare and Early Education Task Data

Egocentric childcare and early education task data is first-person video of one person doing care-room work, recorded from that person’s own point of view. A paid, consented contributor wears the capture rig and performs a task written in a skill spec, and each attempt is delivered as an episode with synchronized streams, hand pose and labeled action segments. It is human capture. It is not robot or teleoperation data, and it is not a fixed corpus you download.

People search for this under several names: childcare robot training data, an early education classroom task video dataset, egocentric childcare worker data or custom care task data collection. They describe one request from different angles. The work is the same: write the task down, collect it on demand, and receive validated episodes.

The catalog places four rows in the Childcare & Early Education environment, which sits in the Education & Childcare sector. The environment examples read: "Daycare center, preschool, nursery -- toy and material organization, feeding, low-angle child interactions, classroom setup." The first table below reproduces the four rows exactly as listed.

The rows are Classroom setup, Feeding, Low-angle child interactions, and Toy and material organization. All four are rated Easy, captured with a head-mounted phone or a multiview rig, and listed as available on request. The second table sets those rows against common childcare requests that are not listed, so a buyer can see where the catalog ends.

Classroom setup reads: "Arrange tables, chairs and learning materials and set up activity stations for the day." Toy and material organization reads: "Sort toys and materials into labelled bins and shelves after activities." Together they are the start and the end of a session, and the spec should say whether each is its own episode or both sit in one.

For both, name the furniture or the item types, the layout or the bin and shelf scheme, and the step boundaries. The site does not state a room type, item count or station layout, so those are yours to set. Two neighbouring pages go deeper on these rows: the classroom setup page covers Classroom setup, and the object rearrangement page covers tidying and putting items away.

Feeding reads: "Prepare and serve meals and snacks to children and help the youngest eat." Low-angle child interactions reads: "Kneel to a child’s height to play, hand over toys and guide activities." Both describe a worker acting with and around children, so the spec has to say which part of the task is the target: the preparation and serving, the handover of a toy, or the whole interaction.

The site does not state a feeding method, a food or utensil type, a child age, a camera height for kneeling or a label for the interaction itself. It also states that depth degrades on wet and reflective surfaces and that a confidence map is delivered so low-confidence regions can be masked; it does not say how food, liquids or low viewpoints change depth quality. Raise any of these in the request and settle them in a Spec Pilot.

Each attempt is delivered as one episode. The site does not set an episode length or count, so the skill spec does. Inside the episode, ten streams run on one hardware clock, and action segments divide the recording into spans, each carrying start_ns, end_ns, verb and noun. A feeding run can become labeled steps such as preparing a plate, serving it and helping someone eat, rather than one long clip.

Two machine-checked rules apply. An episode is rejected if the manipulated object is out of frame for more than 15 percent of it, and if more than 35 percent of depth pixels in the manipulation region are invalid. In childcare work the manipulated object changes often, from a chair to a bin to a spoon to a toy, so the spec should say which object counts as manipulated at each step.

Every contributor signs a written release before capture, and each delivered episode links to that record through consent artefact IDs in the batch data card. The consent chain has six checkpoints, from before capture to after delivery, and a break at any one discards the footage rather than shipping it at a lower grade.

At ingest, automated detection finds faces, plates and screens and blurs them irreversibly. Speech is scrubbed of names, addresses and card numbers while ambient audio is kept. A human reviewer confirms that no un-consented person is identifiable in any frame, and footage that cannot be de-identified is discarded. An undocumented bystander is also one of the machine-checked reject criteria. None of this is legal advice, and the executed agreement controls.

The site states that every participant signs a written release, that bystander faces and identifying details are removed, that any clip where a bystander cannot be de-identified is discarded, and that there are no participants under 18. The person wearing the rig is therefore a consented adult contributor.

The catalog text for Feeding and Low-angle child interactions names children as part of the task. The site does not state whether, how or at what ages a child appears in an episode, and it states no child-specific consent process. This page therefore makes no claim that children are recorded, and none about how a child’s likeness would be handled beyond the process above. A program that needs these two rows should raise that question first, and the answer goes into the written spec and the executed agreement.

The site describes the Childcare & Early Education setting as using a head-mounted phone or a multiview rig. 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. The site says nothing more about the setting beyond the environment name and its examples line.

The site names no daycare, school, district, employer or city, and it lists capture at your own facility as not offered. This page therefore makes no claim that episodes are recorded in a particular facility, and none that they are recorded at yours. Capture is run to your skill spec using the settings and contributors Firsthand arranges.

Delivery is in the format your stack already reads: RLDS, LeRobot, WebDataset, HDF5, zarr or Rerun .rrd, or raw streams with metadata. Readable converter source ships alongside the data so the schema can be retargeted to your own loader. The official documentation for three of these formats is linked in the sources below.

Collections are billed per validated hour, and only footage that clears the signed accept criteria is billed. Rejected items are recaptured at our cost. The license is perpetual, irrevocable and worldwide, buyer-owned and non-exclusive by default, with exclusivity available as a separate term. The site publishes no price list, minimum order or batch size for this environment, so those are set in your quote.

From signed spec to first validated batch, the site states a median of 48 hours where coverage is deep and three to four weeks where it is thin and contributors need to be recruited first. The site does not say whether Childcare & Early Education counts as deep or thin, so ask which timeline applies when you request the quote. Speed comes from a warm pipeline, not from cutting consent, review or anonymization.

The site also states program milestones: a pilot reaches its first delivery in two to three weeks, and a standing program reaches full rate in four to six weeks. After the first batch, on-demand top-ups run against the same accept criteria. The site does not state a turnaround for those later batches, so plan them in your quote.

The classroom setup page covers one row, Classroom setup, and explicitly leaves out Feeding and Low-angle child interactions. The object rearrangement page covers tidying and putting items away as a general task family, which includes the sorting that Toy and material organization describes. This page is the only one that covers all four rows of the environment together, including the two that name children in the task text.

The elder care page covers a different environment and a different row, Feeding assistance, which is a separate catalog entry and not interchangeable with the childcare Feeding row.

Where this site does not state something, this page says so. There is no listed row for naptime, diapering, story time, lesson delivery, outdoor play or end-of-day cleanup. There is no child age, no feeding method, no room or site detail, no site count, no region or city, no episode count and no success or quality label. There is no child-development, educational-outcome or safety-certification claim anywhere on this page, and nothing here reports a robot result, a benchmark or a model improvement.

Counts of contributors, hours and tasks are set in your skill spec. A task outside the four rows is settled in a Spec Pilot rather than assumed.

Describe the task, the items, the setting and the variation or failure cases your pipeline is missing. That becomes a written skill spec, quoted for reach, timeline and price before any capture begins, and most programs run a Spec Pilot first, a small, fast first batch that proves the spec and the accept criteria on real footage before volume scales. A published sample recording is on the sample data outputs page if you want to see the streams first.

Name the step and the item rather than a broad activity such as "childcare," because that is what lets each step become one action segment. If the program needs Feeding or Low-angle child interactions, say so at the start so the footage and privacy questions above are settled in the spec.