Egocentric Home Repair and Handyman Video Data

Egocentric home repair and handyman data is real-world video of a person doing a repair call in a client’s home while wearing the capture rig: diagnosing a fault, fixing a fixture, shelf or minor damage, and testing the result. The footage shows the tool, the fixture and the fault from the head-mounted viewpoint as the person works. Each episode follows the Firsthand episode schema, with ten streams synced under 2 ms. Which streams a head-mounted phone populates, and which the rig adds, is confirmed in the spec.

Teams search for it as a home repair robot training data, handyman task video dataset, egocentric home maintenance data and custom trades task data collection. They all describe demonstrations of a repair call, collected to a written spec. This page covers only what the Firsthand site states and marks everything else as Not stated, including any claim about safety, electrical, gas or code compliance.

The site states four rows in an environment called On-Site Home Repair in its commercial task catalog. "Handyman repair call" reads "Fix doors, shelves, fixtures and minor damage in a client’s home with hand tools" and is rated Medium. "Electrical repair call" reads "Diagnose a fault in a client’s home, replace the failed switch, socket or breaker and test" and is rated Hard. "HVAC service call" reads "Inspect the HVAC unit, replace filters or parts, check refrigerant and test the system" and is rated Hard. "Plumbing repair call" reads "Locate the leak or blockage, replace fittings or seals and test the water flow" and is rated Hard. All four are Available on request.

Each wording names a short arc in order: diagnose, replace or fix, then test. Anything outside that wording is Not stated, including the fixture type, the tool list beyond hand tools, and the fault severity. A policy that needs another repair task writes it into the spec and we scope it, because a catalog row is a starting vocabulary, not a fixed menu.

The Handyman repair call row is rated Medium and sits in the On-Site Home Repair environment, described as "Professional plumber, electrician, HVAC technician, or handyman working in a client’s home". Its capture mode is listed as phone (head-mounted) or multiview rig, and its status is Available on request.

The site defines that status plainly: the task is scoped and recruited to spec when a buyer asks for it. So recording for this row is not already running, and a first batch is quoted against your spec. The row names doors, shelves, fixtures and minor damage, but it states no specific fixture, no damage type and no task duration, and no residence, city or region is named.

The other three rows in the same environment are trade-specific repair calls rather than general handyman work. The electrical row runs from diagnosing a fault to replacing a switch, socket or breaker and testing it. The HVAC row runs from inspecting the unit to replacing filters or parts, checking refrigerant and testing the system. The plumbing row runs from locating a leak or blockage to replacing fittings or seals and testing the water flow.

All three share the environment, the capture modes and the Available on request status with the handyman row, and all three are rated Hard against the handyman row’s Medium. None states a part brand, a code reference or a result, and this page gives no safety, electrical, gas or code-compliance guidance; that is set by the trade’s own standards and the client’s own site rules, not by this page.

Repair-adjacent work appears elsewhere on the site, but on a workshop bench rather than in a home. The Trades Workshop Bench environment lists rows such as "Hang a fixture", described as "Level, mark, and screw-mount a fixture", "Apply + tool caulk", described as "Run a caulk bead and tool it smooth", and "Drive screws with driver", described as "Drive screws with a power driver". All three are rated Medium, Available on request, and captured with a multiview rig plus handheld grippers rather than the phone or rig modes used in homes.

Those bench rows share tools and movements with the handyman row but are not stated as happening inside a residence. This page covers only the four On-Site Home Repair rows.

On-Site Home Repair is described as "Professional plumber, electrician, HVAC technician, or handyman working in a client’s home". It holds four rows, all Available on request, and all listing a head-mounted phone or a multiview rig.

So the site describes work performed by a professional tradesperson inside a client’s residence, in the Construction & Trades sector, and not a staged set. It defines the phone mode as a head-mounted phone recording the first-person view of real work on site, and the rig mode as adding synchronized third-person angles when the spec asks for them. The rows do not say which mode applies to which row.

The site states that egocentric capture happens in real kitchens, workplaces and streets. It names no city, region, client or address, and does not state the number of residences, sites or episodes. Environment, access and schedule are agreed in the written spec, and whose home is used, and under what access agreement, is set per program rather than stated here.

In the phone mode, a head-mounted phone records first-person video as the person works through the repair call. In the rig mode, a multiview rig adds synchronized angles when the spec asks for them. Camera count, resolution, frame rate, mounting position and tool list are Not stated for these rows beyond "hand tools" on the handyman row. Which streams each mode populates is confirmed in the spec.

Each episode follows the Firsthand episode schema: ten streams synced under 2 ms, with RGB, depth, 3D hand pose with per-joint visibility, contact and grasp events, and frame-accurate action segments. The site also states a 200 Hz 6-DoF IMU among its sensors. For home repair, action segments are where a spec marks each step: the diagnosis, the fix or replacement, and the test. Contact events mark when a hand meets a tool or a fixture.

Not stated: whether depth and hand pose are populated in the phone mode, and whether a fault, fixture or test result is labeled as a success or quality field. This page makes no claim about them. Task duration, the number of episodes, residences or sites, and robot results are Not stated too.

The site gives no technical analysis of home repair, so this page makes no claim about robot results. It does say, on its mobile manipulation page, that a robot working in a home or a facility does not stay at one bench: it carries things between rooms and crosses a floor to the next station, and a repair call moves between the entry, the fault location and wherever the tools and parts are kept.

What the row wording implies is a different data shape from bench assembly: the fault is diagnosed before it is fixed, which the rows capture as an ordered arc rather than a single grasp. The site rates three of the four rows Hard and one Medium, which is the only statement on difficulty. It states no success or quality label for these rows, so this page makes no claim that any row demonstrates a repair done correctly, only that the row names the arc a spec can capture.

Delivery formats are RLDS, LeRobot and WebDataset, plus HDF5, zarr and Rerun .rrd. Home repair episodes use the same schema as every other Firsthand collection, so a loader that reads another Firsthand dataset reads this one without changes. The format table below sets each format 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. Rejected items are recaptured at our cost, and billing is per validated hour.

Every episode links to a signed contributor release through a documented consent chain, and the consent record is part of the delivery. The release also covers bystanders: contributors confirm they have not knowingly recorded anyone who has not signed it, and where an unidentified person appears, the footage is de-identified or discarded. Capture happens inside a client’s home, so site and resident access is a condition the spec sets, not a detail this page states.

A vision model runs on every frame at ingest, flagging faces, heads, licence plates and any legible screen or document. Those regions are irreversibly blurred, and a human reviewer confirms the result before footage ships. If the reviewer is unsure a region is fully covered, the episode is held or discarded, never downgraded and shipped. The site does not state a specific rule for residents or bystanders inside the home beyond the general consent chain, so raise any such need before the spec is signed.

Where coverage is already deep, the first validated batch lands in a median of 48 hours from signed spec. Thin environments typically take three to four weeks, because vetting operators there is the bottleneck, and we quote that up front. All four On-Site Home Repair rows are Available on request, so they are recruited to your spec first, and the 48 hour figure should not be assumed.

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. Those assume capture coverage is already in place. A batch size and a later-batch turnaround are Not stated, so both are quoted against your spec.

You describe the repair calls, the trades and the steps your policy needs. We turn that into a written spec covering tasks, environments, capture tier, annotation depth and accept criteria, then quote reach, timeline and price before any capture begins. This is the custom egocentric data collection route that every Firsthand task family follows.

A Spec Pilot, a small and fast first batch, can prove the spec and accept criteria before you scale. Every collection ships under a perpetual, buyer-owned commercial license, with a reject log and signed accept criteria on every batch.