Stereo Furniture Assembly Data

It is first-person video of a person doing the joining part of furniture work: gluing and clamping frame joints, drilling pilot holes and screwing the furniture frame together, or assembling commercial furniture units from components with hand and power tools. It is recorded from their own point of view with two calibrated cameras, with hand pose and sensor streams alongside. Stereo describes the capture, and MCAP describes the container. They are separate choices.

The site does not state that stereo is already running on these rows. The catalog lists a head-mounted phone with a multiview rig as their capture setup, so a stereo pair is a capture choice agreed in the skill spec and confirmed on a Spec Pilot. 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. Firsthand is not affiliated with Foxglove or the MCAP maintainers.

Two, both in the Carpentry & Furniture Workshop environment of the Manufacturing & Workshops sector. "Assemble and join furniture frames using wood glue, clamps, and a drill" is described as "Glue and clamp frame joints, drill pilot holes and screw the furniture frame together." "Commercial Furniture Assembly" is described as "Assembling commercial furniture units from components using hand and power tools." Both are rated Medium.

Both are listed as Active collection. The catalog does not define that status further, and the site says nothing about delivered hours or delivered stereo footage for either row, so this page claims none. The only tools the first row names are wood glue, clamps and a drill, and the second names hand and power tools without naming any. The first table below reproduces the rows as the catalog states them.

The egocentric furniture assembly page covers these two rows and the "Assemble flat-pack" row as first-person catalog capture, with no stereo or MCAP treatment. The stereo table saw and wood cutting page covers two cutting rows in the same workshop environment, both Hard and Active collection. The stereo carpentry and framing page covers five carpentry rows across two environments. The stereo consumer goods assembly page covers three rows in a different environment, all Ready to deploy.

This page is the stereo and MCAP treatment of the two furniture assembly rows only, with its own stream map and failure modes. The second table below sets the pages side by side.

Frame joining is close, cluttered work. A stereo pair recovers depth from the calibrated frame pair rather than from a standalone depth sensor, so a frame part, the clamp that holds it and the hand around both share a frame. The site gives no depth accuracy figure for these rows and no label for joint fit, glue squeeze-out, assembly completion or success, so what a buyer does with that depth is their pipeline question.

A stereo pair adds depth, not a verdict. The site states 1080p per eye for dataset 06, and whether that suffices to see a pilot hole or a screw head at working distance is Not stated, so a spec should check it on a Spec Pilot. The site also states that Firsthand does not capture at a customer own facility.

Joining a frame is two-handed: one hand steadies or closes a clamp while the other places a part or drives the drill. The site states 3D hand pose of 21 joints per hand with a per-joint visibility flag, plus contact and grasp events that mark when fingers meet an object, so each hand has its own record.

The site does not say whether a frame part, a clamp or the drill counts as the object in a contact event, so settle what an event means for your task in the spec. Wrist cameras give a closer view of the hands, and a multiview rig adds synchronized third-person angles when the spec asks for them.

Clamps and frame parts hide fingers, and the head camera sits close to the work, as close as a robot onboard camera would. The site states that fingers disappear behind panels and tools, which is why each hand joint carries a visibility flag, and that small parts can fall below a depth sensor reliable floor: hand pose stays valid at that range, and metric depth on sub-centimetre parts is flagged rather than interpolated.

Whether a screw head or a pilot hole counts as such a part is Not stated, and the site gives no minimum object size and no depth figure at working distance. The site also states that depth degrades on wet and reflective surfaces and ships per-frame confidence maps so a buyer can mask rather than guess. How depth behaves on wet wood glue, a clamp surface or a drill body in particular is Not stated.

Tool contact is stated only as contact events that mark when a hand meets an object, and as action segments with a verb and a noun. The site states no force, torque or tactile reading, and it states that the catalog rows name only wood glue, clamps and a drill for the first row and hand and power tools for the second.

The catalog rows state no tool-safety rule, and this page states none. Drill models, clamp types, adhesive brands and torque settings are Not stated, so any requirement belongs in the written spec.

The site describes two stereo setups and does not say which one a furniture assembly order uses. The hard-environments stack is a global-shutter 4K/60 head camera with a 120° lens, two 1080p/60 wrist cameras, a depth sensor at 848×480 and 30 fps, and 200 Hz 6-DoF IMUs. The off-the-shelf stereo head rig, dataset 06, is dual fisheye cameras at 1080p per eye, a 62 mm baseline, per-unit calibration, a 9-axis IMU and hand-tracking output.

Which rig, resolution and streams an order uses is Not stated on the site, and the third table below sets what is stated beside what is not. Intrinsics and extrinsics are re-solved at the start and end of every session, and streams are hardware-triggered onto one clock with a 2.0 ms cross-stream ceiling.

MCAP is an open source container file format for multimodal log data, supporting multiple channels of timestamped pre-serialized data. It is serialization-agnostic, so messages can be Protobuf, DDS (CDR), ROS, JSON and more. Its design is row-oriented and append-only, it stores schemas alongside the data, and it offers an optional index plus optional LZ4 or Zstandard chunk compression.

Reader and writer libraries exist for C++, Go, Python, Rust, Swift and TypeScript, and the mcap CLI can inspect, merge and split files. The foxglove/mcap repository is MIT-licensed; that is the MCAP project licence, not the licence of a Firsthand delivery. The MCAP hub page and the glossary entry cover the format in full.

Firsthand offers MCAP delivery, and dataset 06 is stated as shipping as MCAP. For these rows the site states hand pose, a stereo-derived depth stream with confidence maps, IMU, contact events, action segments, per-frame instance masks and one task instruction per episode elsewhere. What it does not state is which of those are channels in an MCAP file, how they are named, what encoding they use or at what rate.

No topic name, message schema or rate is stated on this page, and none should be assumed. Name the streams that matter in the skill spec and confirm them on a Spec Pilot batch. The fourth table below sets each stream beside what the site states about it.

Hand occlusion is the first risk. Fingers disappear behind a frame part, a clamp or the drill, which is why each hand joint carries a visibility flag, and a spec can ask for wrist views or a second angle. An episode is rejected when the manipulated object is out of frame for more than 15 percent of it. For furniture the spec decides what that object is: the frame part, the clamp, the drill or the unit being assembled.

Invalid depth near the work is the second. An episode is rejected when more than 35 percent of depth pixels in the manipulation region are invalid. Calibration is measured at the start and end of every session under a 0.28 px RMS gate, and an episode is rejected for a sync breach beyond the 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.

Furniture types, brands, part counts, cycle times, tolerances, workshop or customer names and safety certifications are Not stated and are not claimed here. The fifth table below lists the failure modes beside what the site states.

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 an undocumented bystander is a stated reject criterion.

The site does not state a rule specific to a workshop, or to branded furniture parts or packaging, so this page makes no such claim. Settle it in the spec.

The October 2026 off-the-shelf catalog lists ~324,800 hours across 18 datasets, of which stereo is ~64,800 hours across 9 datasets. No listed dataset names furniture assembly, furniture frame joining or clamping with wood glue, and the site does not state that any contains either row.

Stereo furniture assembly in MCAP is therefore custom collection, not an off-the-shelf purchase: 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. The site does not say which speed applies to these rows, so the timeline is quoted against your spec. Collections are billed per validated hour, and rejected episodes and recaptures are never billed.

Describe the steps and where each starts and ends: gluing a joint, closing a clamp, drilling a pilot hole, screwing the frame, or assembling a unit from components. Say whether an episode covers one joint or a whole frame, since the site does not set an episode length, and name the variation your pipeline is missing. Then name the streams you want in the file, the topic naming and message encoding your reader expects, whether calibration ships inside the file or beside it, and whether MCAP is the only delivery or sits beside RLDS, LeRobot or WebDataset.

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 sixth table below shows what each spec element decides.