Signing on delivery day is easy; reconciling the data three months later is not. Twelve checks in three stages: before unboxing, on installation day, and across the first week.
Acceptance is the step most often skipped on delivery day. Sales is present, the engineer is present, the department head is present, and all three want the device working that afternoon — so signing becomes a formality. Three months later the data does not reconcile, report fields are missing and training never happened, and by then the device is already inside the workflow.
Acceptance is not fault-finding. It is writing down, before signature, what this device can and cannot do. The twelve checks below run in three stages: three before unboxing, six on installation day, three across the first week. Each has a method, a pass criterion and a remedy.
1. Before unboxing: verify three documents
Once the device is unboxed, powered and networked, the boundary for return becomes unclear. So check the paperwork before breaking the seal.
Contract and configuration list. Model, quantity and accessories item by item. Pay attention to the calibration artefact, positioning marks and power adapter. Without the calibration artefact, the accuracy check below cannot be done.
Registration certificate and product composition. A medical-device edition must have a registration certificate, and the composition, model and structure listed on its annex must match the physical unit. A different suffix is a different product.
Accompanying documents. Manual, certificate of conformity, warranty card, software licence terms. The licence matters especially: perpetual with the device, or annual subscription — that determines the cost structure three years out.
If the three do not agree, do not unbox. Ask the supplier for a written explanation first. Once the box is open, your negotiating position changes.
2. Installation day: measure six dimensions
Many "it feels awkward" problems are not device problems but site problems — and they are fixed the moment the device lands. Measure six dimensions and write them on the acceptance form.
Standing depth. Distance from lens to standing position, plus clearance to the wall behind. Too close and the body is cropped; too far and point density drops.
Standing width. Arms hanging naturally stay inside the capture volume, and a guide can stand alongside.
Clear height. Including clearance from the top of the device to light fittings and sprinklers. Enclosed units such as a body-measuring booth have hard height requirements.
Passage width. Whether a stretcher, wheelchair or pram can pass. Mandatory for maternal and child settings and community health centres.
Power position and earthing. Distance from socket to device, and whether an independent earth exists. Sharing a power strip with high-draw equipment introduces interference.
Network and signal. Whether wired access is available and how strong the wireless signal is. QR-code report retrieval depends on it; a weak signal looks like "the report will not open" and is easily misdiagnosed as a software fault.
Record all six on the spot. If the standing position later proves unsuitable, the fix is on the site, not the device — and this record is what makes the responsibility clear.
3. Accuracy: one person, three captures, then a changed posture
Millimetre-level is a nominal figure with preconditions: close-fitting clothing, bare feet, natural stance, still for three to five seconds, hair tied low. Acceptance is not about reproducing the factory lab; it is about reproducing results in your site, with your staff, in your process.
Repeatability. Capture a colleague three times under standard conditions and compare the same dimension. A large spread points to inconsistent positioning or timing — check the process before the device.
Sensitivity. Change the posture once, for example a slight forward lean, and capture again. The data must move visibly. If it barely changes, the reconstruction is guessing and the report's credibility is questionable.
Boundaries. Capture once in a thick coat, once with a high ponytail, once without bare feet, and record the deviation. These are the everyday non-standard conditions; knowing the deviation is what lets you write "for reference only" on a report.
Note: values obtained outside standard conditions from a living subject are only suitable for initial screening and risk indication, not as measurement conclusions. Put this in the department's operating standard, not just on the acceptance form.
4. Software and report: check fields, not the interface
Field completeness. Every item on the report must map to a field in the database with an explicit unit. A report that looks good but cannot export data is effectively no data.
Export. Structured export (table or API) and 3D model export. Test the format and units on the spot.
Permissions and audit. Who can view, who can edit, who exported what, and whether it is logged. This is a hard requirement where patient data is involved.
Historical comparison. Two captures of the same person displayed side by side with the difference. Without it, follow-up degenerates into "measure again".
5. Training: the operator runs it alone
Training is not listening once, it is doing once. Have the person who will actually operate the device complete a full cycle unaided: start-up self-check, capture, report, briefing, record creation, export. The acceptor watches and does not prompt.
Three criteria: completes it alone, can explain what each item means, and knows when to recommend medical attention. The third is most often skipped and matters most for compliance — turning a number on a report into a diagnosis puts the risk on the institution, not the device.
6. The first week: watch three things
Capture failure rate. Log daily failures and re-tests with reasons. A sudden rise points first to ambient light and positioning, then to the device.
Report open rate. The proportion retrieved by QR code. A low rate is usually network or entry-point design, not lack of interest.
Record completeness. How many visits created a record, and how many produced only a report. Without complete records, follow-up and repeat purchase have nothing to stand on.
7. When something fails: three categories
Document failures — model mismatch, missing certificate annex, unclear software licence. Do not sign, do not use, set a deadline to supply. These cannot be fixed later.
Performance failures — poor repeatability, export failure, missing report fields. Set a rectification deadline, record the affected scope, and re-test the same way afterwards.
Site failures — insufficient standing space, light interference, unstable network. These are the institution's to fix; the supplier gives written recommendations and the institution adjusts and re-tests.
Separating the three keeps the conversation from becoming mutual blame. Add a "outstanding items and rectification deadline" section to the signature page and keep a copy each — it is the only evidence three months later.
8. After acceptance: turn the form into an operating standard
Filing the acceptance form and forgetting it wastes it. At least four of its contents should become daily operating rules: standing mark, capture conditions, re-test method, exception handling.
Standing mark. The position fixed during acceptance is the position used daily. One sentence taped next to the device beats one mention during training.
Capture conditions. Close-fitting clothing, bare feet, several seconds still — these are the preconditions under which the report holds. Put them in the report footer so readers know too.
Re-test method. A re-test reuses the original conditions; that is what makes follow-up meaningful.
Exception handling. What to self-check, what to escalate to the manufacturer, what to stop using. Three categories, so front-line staff feel able to decide.
Four sentences on the wall get followed more often than a thick manual. Whether a device is used well often comes down to whether these few sentences are visible.
Next step
If you want the twelve checks adapted to your setting, send your institution type and the models under consideration. Cooperation runs as direct purchase or through a dealer arrangement; details are discussed directly.
Note on figures: nominal accuracy is the manufacturer's claim and only holds under the stated conditions. Acceptance methods and criteria are implementation recommendations. This article contains no pricing.
Measurement conditions and what the result covers
The accuracy figures and posture data discussed here all rest on the same measurement conditions: the person wears close-fitting clothing or shorts, is barefoot, stands naturally, stays as still as possible during the scan (about 3–5 seconds), and has hair tied back or pinned up so it does not cover the head and shoulder outline.
When these conditions are met, the data is used for preliminary screening, risk indication and follow-up comparison of posture and spinal status; when they are not met (loose clothing, movement, hair covering the outline), the result is an estimate for reference only. The device does not provide a medical diagnosis — any abnormal finding needs review at a medical institution.
Accuracy is millimetre-level (vendor nominal ±1 mm). A living body is not a calibration artefact: millimetre-level variation across three consecutive scans of the same person is normal, so read the trend from the range of several measurements rather than from a single number.
Where the data comes from
Specifications here come from the device datasheet and user manual — nothing is estimated or inferred.
Case data comes from client project records; organization names can be verified by phone. We do not use client names without authorization.
Qualification records can be looked up by certificate number in the National Medical Products Administration database.
FAQ
What is most often missed at acceptance?
Whether report fields can be exported. Most people check that a report can be produced, not that the data can be taken out, and discover at the point of departmental reporting or research analysis that they have images but no structured data. Export once on acceptance day; it takes five minutes.
Is one subject enough for the accuracy check?
No. Cover at least two body types: one standard and one outside the norm, for example unusually tall or heavy. The standard subject tests repeatability; the outlier tests the boundary. Performance at the edge of the body-type range is where everyday problems actually appear.
The supplier says it works on installation day. Is a first week of observation still needed?
Yes. Working on installation day does not mean working a week later: light changes with the season, the network changes with load, and operators change with familiarity. Watch three numbers in the first week: capture failure rate, report open rate and record completeness.
What single line must appear on the acceptance form?
Outstanding issues and their correction deadline. Acceptance is almost never all green, and anything not written down becomes a dispute three months later. Keep a column on the signature page with one copy for each side; it is the only record that will hold up afterwards.