Asset Checks, Label Governance and the Risk Score

An asset check asks custodians to confirm, asset by asset, that they still hold what the register says. Open Assets, then Asset Checks, and create a campaign. Its scope can be narrowed by category, warehouse, location, custodian, the risk level you set on the asset, or the calculated risk score, either by band or by a minimum score. Only individually tracked assets that are with a custodian and are not lost or retired are taken in, because there is nobody to ask about pooled stock or an asset sitting in a store. Preview tells you how many assets the scope takes in before you save it.

Each campaign has a due date, a number of days after it before escalation, and the evidence every answer must carry: a label scan that must read this asset’s label, a GPS position with an optional accuracy limit in metres, and a photo. A campaign starts as a draft. Launch takes a snapshot of every asset in scope with the custodian holding it at that moment, one row each, and from then on only the instructions and dates can change, never the list. Close the campaign when the sweep is done, or cancel it.

In practice: an NGO in Kisumu has 140 laptops and 30 motorbikes with field staff. The operations manager scopes a campaign to the IT Equipment and Vehicles categories, requires a label scan and GPS within 50 metres, sets a due date ten days out and escalation after three days, previews 170 assets, and launches. A motorbike reassigned the following day has changed hands since launch, so its row can no longer be answered by anyone; the manager waives it with the reason, and the new rider answers for it in the next campaign.

Custodians answer under My Asset Checks on the web or in the AWRA app. Confirmed means they have the asset in hand, and it records a verified movement, so the asset’s last-verified date and its movement history say the same thing. Not held and Damaged are exceptions for a manager and need a note saying where the asset is or what the damage is: answering never moves the asset, because marking an asset lost is a movement with its own permission. A manager with the custodian and the asset in front of them can record a witnessed answer, and a row that should not be in the campaign can be waived with a reason.

Contractors and casual staff often hold assets without an AWRA account. Their reminder email carries a signed link to a page where they answer for their own assets under the same scan, GPS and photo rules. The link lasts until 30 days after the due date and never less than seven days from when it was sent, and a tampered or expired link simply shows as not found. Reminders go out when the campaign opens, two days before it is due, and the day after if the custodian still has not answered, each one listing all their open assets.

Once the escalation days have passed, each custodian’s line manager, taken from the employee record, is told who has not answered, and again weekly while the campaign is open. People who manage asset checks hear about Not held and Damaged answers the morning after. In practice: a site supervisor answers Not held for a generator. Nothing moves; the asset manager sees the exception next morning, finds the generator was lent to another site, records the transfer properly, and only then decides whether anything is actually lost.

When anybody can print a label at any time, a worn label and a label moved onto another laptop look the same. AWRA records labels instead. A first label is queued from the asset list or the asset page, and queuing the same asset twice gives back the label already queued. On the Label Queue, queued labels are gathered into a batch that keeps its print layout. Print the batch PDF and press Mark printed: only then do those labels become the active ones. If the print failed, cancel the batch and its labels return to the queue.

An asset that already has an active label gets another only through a replacement request with a reason, and an asset can have only one open request at a time. Somebody other than the requester approves or rejects it, and a rejection needs a reason the requester will see. Approval puts the replacement in the queue, and the old label stays active until the new one is marked printed, so the asset is never without a valid label. Reporting a label as unreadable, damaged, tampered or missing, on the web or in the app, logs the problem and opens a replacement request automatically.

Every successful scan is counted on the active label, which lets the Label Queue show health: replacement pending, not scanning, waiting to print, no tracked label, never scanned, not scanned recently (180 days) or scanning normally. In practice: a hospital’s biomedical team reports a tampered tag on an infusion pump. The report opens a replacement request, the asset manager rejects a second request from the same technician as a duplicate, approves the first, prints it with that afternoon’s batch, and the old tag stops being the active label the moment the new one is marked printed.

Each asset carries a score from 0 to 100, built from six parts that each have a ceiling: book value against your workspace reference (up to 25), category and the risk level you set (15), how many distinct places it has been in 180 days (15), overdue custody, growing with the days (20), how often it moved in 90 days (10), and verification (15), which counts an exception, an overdue answer, or a last verification that is missing or older than 180 days. The score falls into a band, and the asset page shows how it was reached.

The score is recalculated after a movement, after an asset check answer, when a scoring field on the asset changes, and every night, because an overdue return gets riskier even when nothing happens. Your own risk level is never overwritten; it is one of the inputs. A score is advisory: if it cannot be calculated, the movement or answer that asked for it still goes through.

In practice: an asset manager filters the register by the high and critical bands, finds that most are laptops out with staff past their return dates, and launches a campaign scoped to a minimum score of 60. After the campaign, the confirmed assets drop in score because their verification is now recent, and the remaining high scores point at the handful that need a conversation rather than a sweep.