Vendor access
This part is for Aigna staff
Vendors see their own installation and the device reference they need to build it: device types, setup guides and models. Everything else stays with Aigna.
Go to your installationKeia Method· Level 2 · 2.2 · idea 2 of 5
Capture, don't ask
Fill in every record from what is already known, and ask people only what only they know.
Capture, don't ask
- Instead of
- forms that ask people what you already know.
- You
- fill in each new job from your records (room, device, last change) and ask people only what only they know.
- You'll know it's working when
- a report needs no follow-up question, and closing a job takes one sentence: what fixed it.
In Keia Atlas: an alert opens a job already filled in; a QR code on the table reports a fault in one tap.
In the morning story, this is 07:53: A job opens with those facts filled in, and the last change beside them: the switch on that floor was updated on Tuesday. Nobody is asked for a room number or an asset tag. With a spreadsheet, the desk copies the same facts from the room's row. Read the story
In full, with no software
- Instead of
- Forms that ask what the alert already said. The portal is the slowest way to report a fault: 4 h 26 min lost per incident, against 1 h 24 min for a walk-in (HappySignals, 2026). Only 44% of issues are reported at all (Nexthink, 2023).
- You
- Fill every record from what is known before a person sees it: the alert names the unit, the unit knows its space, the calendar knows the booking, the build sheet (every setting of one unit) knows the settings, and the change log knows what happened last. The categories are the space and the device type the record already has. Ask a person one thing at a time, and only what only they know.
- The moves
- Confirm accept what was captured, correct what was not
- Capture one sentence at the moment of knowing: "What fixed it?" or "Nothing new"
- Propose a captured fix that keeps helping becomes a proposed edit to the setup guide, approved by its owner
- You'll know it's working when
- More fields are filled by capture than typed; reports per space come close to the faults you observe; the time from a fault to a filled-in incident falls. Measure all three.
In Keia Atlas
Where each field came from
Alert · 07:52
dub-309-vc01 offline
fleet tool
Incident · opened 07:52 · filled in before anyone saw it
- Space
- 3.09 Whooper Swan, Dublin officefrom the space record
- Unit
- Video bar, Poly Studio X72 (dub-309-vc01)from the fleet tool
- Resembles
- A known error for this modelfrom the known errors
- Last change
- Firmware, twelve days agofrom the change log
- Priority
- Proposed from the 09:00 booking; a person may change itfrom the calendar
- What fixed it?
- Asked once, when it closesfrom you
An incident arrives with its space, unit, likely known error, last change and a proposed priority a person may change. The room guide, the page behind the code on each meeting-room table, reports a fault in one tap. A change is written from the work, not typed after it. A project's as-built record becomes the operational record the next day.
Why it works, and its limits
- Why it works
- The form is the slowest, least-liked way to report, and half of problems are never reported. Extraneous load (hunting, re-typing, translating) is the only kind of mental load design can remove.
- Limits
- Capture needs feeds. Much works with no connector (the catalogue, space types, floor plans, standards, build sheets, the room guide); every connected figure is labelled with its source and time.
What it maps to
| On screen | Plain meaning | Maps to |
|---|---|---|
| Incident, Major incident | Something is broken; one so big it runs the urgency rule and names a commander | ITIL incident |
| Priority P1 to P4 | Proposed from impact and urgency; a person may change it, with the reason kept | ITIL priority |
| Setup guide, Build sheet | The settings for a model, in order; every setting's actual value for one unit | Configuration |
| Problem, Known error | The cause behind repeating incidents; a cause with a known workaround | ITIL problem, known error |
Try it in the demo
- Open the room guide for 3.09 Whooper Swan, the page behind the code on the table.
- Report a fault: tap the symptom. The space and its devices are already known, so nothing else is asked.