Quality Inspection for Utility Work Orders

Quality inspection for utility work orders answers one question before any money is claimed: was the work done as required? The answer comes not from a supervisor’s impression but from defined evaluation items, checked one by one, with a note and evidence behind every verdict. Pass every item and the order moves on to the achievement certificate; fail one and it goes back to execution.

Written for quality managers at electricity-network contractors, this guide covers how to design evaluation items, what evidence to require, how the return-and-response loop runs to close-out, which indicators to track, and how the cycle works in ELX.

Where does quality inspection for utility work orders fit?

A work order typically passes through the site survey (Kashfia), an excavation permit where needed, and execution in Projects or Connections, then reaches quality, the achievement certificate, drawing and the extract. Quality is the last operational checkpoint before completed work becomes a value you can claim. The full journey is in our guide to utility work order stages.

Two consequences follow. A defect that slips past quality travels into the certificate and the extract, where it is far harder to fix. And a slow quality stage delays everything after it, since certificates wait for quality sign-off and extracts are built from certificates.

Evaluation items

The checks executed work is measured against, each with one verdict: identical (pass) or not identical (fail).

Identical / not identical

The verdict on an item; “not identical” needs a note pinning down the defect and evidence to back it.

Response to remarks

Files and notes the execution team attaches to prove a fix before resubmitting to quality.

Internal quality versus regulatory requirements

The Ministry of Municipalities and Housing (MOMAH) describes Balady’s building-connection excavation permit as a path from coordination and approval to permit issue, then the excavation work and, finally, closing the excavation. Its conditions include attaching the work order to the request and preparing a site safety and security plan.

MOMAH’s requirements for a new excavation permit also state that using subcontractors does not exempt the main contractor from liability for violations. That makes a strong case for an internal quality check that holds the main contractor’s own crews and its subcontractors to one standard, and gathers evidence during the work, before the trench is backfilled and the excavation closed.

That check does not replace regulatory requirements or site safety rules, which we cover in a separate guide to excavation safety requirements. The question here is whether executed work conforms to what was required.

How do you design quality evaluation items?

Vague items let two inspectors reach different verdicts on the same order; precise, verifiable ones give the same verdict whoever inspects. Five design principles:

• One item, one check. Don’t combine trench depth and site cleanliness; if one passes and the other fails, the verdict means nothing.
• Verifiable wording. Write items any inspector can confirm on site or from photos: “executed excavation dimensions match the execution record”, not “work is acceptable”.
• Binary verdicts. Identical or not identical. Scores invite judgement calls; a binary verdict makes the return decision obvious.
• A reference for every item. Tie each item to its source: the contract terms, the technical specifications they name, or the approved drawing.
• One short, shared list. Define items once, in a central guide every team uses. Add an item when a defect the list misses keeps recurring, and periodically review items that have never failed: do they still earn their place?

Which evidence should be attached to each item?

A verdict without evidence is an opinion, and opinions get disputed. Every not-identical item needs a note describing the defect and a file proving it; items that cannot be verified later deserve evidence even when they pass.

Capture evidence at the right moment: some items cannot be checked after backfill. Give each file a reference name tied to its item, such as “Remark 2 – path depth”. And write notes that stand on their own: what is wrong, where, and what the fix requires.

What happens when an item is not identical?

One failed item sends the whole work order back to execution. The logic is financial: the achievement certificate is built from executed items and quantities, and a fix can change a quantity or add work, so a certificate prepared before every remark is closed may have to be recalculated after the fix.

Hence the all-or-nothing rule. The return loop usually runs in five steps:

1. Verdict and remark

   The inspector marks each failed item with a note and evidence.

2. Return to execution

   The order goes back to the executing team in Projects or Connections and stays with them until the remarks are resolved.

3. Fix and respond

   The team resolves each remark, attaches post-fix files and notes, and resubmits.

4. Re-inspect

   The inspector re-checks the returned items and any item the fix may have affected.

5. Close

   When every item passes, the order moves on to the achievement certificate.

Each return adds days to the order’s age and delays its extract, so the return count deserves tracking as a KPI. For what comes next, read achievement certificate vs. extract.

How does the execution team respond to quality remarks?

A good response closes a remark the first time; a weak one opens another return cycle. Often the difference lies in how the response is documented, not in the fix itself:

• Answer each remark individually. “Fixed” for the whole order is not enough; name each item and what was done.
• Before and after. Take the post-fix photo from the same angle as the remark photo, under a matching reference name.
• No evidence, no resubmission. Resubmitting without files usually means another return.
• Update execution data. If the fix changed a quantity or excavation dimensions, correct the execution record first so the certificate starts from the right figures.
• Dispute with evidence. If a remark seems wrong, say why and attach proof instead of ignoring it.

Which indicators measure quality performance?

Passing quality is not the whole story; what matters is how long it took and how often the order came back. Five indicators give a clear view:

Compare across teams, subcontractors and work types rather than reading raw numbers alone; complex jobs may come back more often for good reasons. Treat these indicators as a management framework for your quality team, not a list of ready-made reports; the next section shows what ELX provides.

For stage time limits and delay justifications, see our guide to work order SLAs; for stage indicators on one screen, see the work order dashboard guide.

How quality is managed in ELX

ELX is a bilingual cloud (SaaS) service, hosted and operated by Matrix, that runs in the browser with no servers needed on the contractor’s side. It is built for electricity-network contractors in Saudi Arabia to manage each work order from registration to archive. Quality is a stage of its own within execution and quality control. It follows Projects or Connections (or the Replay stage, when the project includes a drill type flagged “Replay screen”) and precedes the achievement certificate.

Item-by-item evaluation

• Items from a guide. Quality evaluation items are master data with a name per language. They are added, edited, activated or deactivated from the guide screen, and changes are recorded in the audit log. Guides sit under permissions, settings and master data.
• A verdict per item. Identical or not identical, with notes and files for that specific item.
• Execution context. The quality screen shows the excavations and implementation items recorded in execution plus files from earlier stages, and lets the inspector upload quality files.

From return to close-out

• Automatic return. If any item is not identical, the order returns to Projects or Connections and ELX counts the returns; the execution team replies with files and notes and resends it to quality.
• The 95% / 100% rule. During execution, completion is capped at 95% and the done-work value is calculated automatically from the estimated value. When every item passes, completion moves to 100%, the done-work value to the estimate, and the order goes to the achievement certificate in extracts and achievement certificates.

On the normal path, then, an order cannot reach the achievement certificate until every item passes, and its completion cannot reach 100% before quality signs it off. Moving an order to another stage manually is limited to authorised users, requires a mandatory note and is logged separately with the old value, new value and date.

Distribution, time limits and follow-up

• Assignment. An order reaches the quality list directly or via a distribution basket, where a dispatcher assigns it to a user permitted on the quality screen for the order’s district. The assignee gets a real-time in-app alert and a browser notification, then claims the order with “Open”.
• Stage time limit. Administrators set the limit in days and can make notes, attachments or a delay justification (once the limit is passed) mandatory before relay.
• Dashboard and alerts. The dashboard shows quality orders within and past the limit, and clicking the late slice lists those orders and who holds each one (see reports and dashboards). When an order enters a new stage, a Telegram alert (and a WhatsApp alert, when enabled) is sent to users who can view that stage’s screen and have notifications on.
• Complete file. The work order lifecycle report includes a quality section and can be downloaded as a PDF or printed; the attachments screen downloads every stage’s files as a single ZIP.

Want to see the ELX quality loop from evaluation to close-out? Request a demo. If your contracts call for a quality screen or report of their own, ask about customization: Matrix tailors screens, reports and stages to each contractor’s needs.

Checklist before relaying an order out of quality

• Every evaluation item has a recorded verdict: identical or not identical.
• Every previously failed item has a documented response: a file and a note.
• Evidence is uploaded for items that cannot be verified after backfill.
• Execution data reflects any quantity or dimension a fix changed.
• No remarks remain open before the achievement certificate.
• A delay justification is entered if the stage time limit was exceeded.