An older bridge-inspection article might explain crack mapping well but give outdated advice on recording findings. Replacing the whole piece would lose a useful explanation; changing only its publication date would leave the bad advice in place. Start by separating what remains technically valid from what has changed.
That distinction matters in transport infrastructure publishing. Engineers may turn to an article to interpret field observations, compare investigation methods, or frame questions for a project team. A revision should help them do that without implying that general guidance can replace site-specific investigation, current requirements, or engineering judgment.
Decide what kind of work the article needs
Not every old page needs a full rewrite. Identify what it set out to explain and why it needs attention now. An explanation of how groundwater pressure affects tunnel lining loads has a different shelf life from a description of a particular monitoring product or a time-sensitive recovery procedure.
- Correct: Fix a discrete factual error, broken figure, mislabeled unit, or statement that no longer reflects the cited evidence.
- Update: Add relevant developments while keeping the sound explanation and structure.
- Reframe: Change the scope if the original question is too broad, poorly defined, or no longer useful to the intended reader.
- Retire or consolidate: Remove a page from active use if its premise cannot be supported or another article now answers the same question more accurately.
Age alone does not settle the choice. A ten-year-old explanation of effective stress may need little work; a recent account of an inspection workflow may already be incomplete. Prioritize pages that could lead readers to a consequential wrong inference, rather than simply working through the oldest dates first.
Audit claims before editing prose
Read the article claim by claim. Mark statements about mechanisms, observed performance, recommended practices, numerical values, standards, and limitations. They need different checks. A physical mechanism may still hold even if an example project has changed. A numerical threshold or regulatory reference can mislead if its jurisdiction, edition, test method, or assumptions are unclear.
A short audit record makes the decisions traceable:
| Record | Question to resolve | Typical action |
|---|---|---|
| Core explanation | Does the mechanism still hold under the stated conditions? | Retain, clarify, or qualify |
| Evidence | Can the source and its scope still be verified? | Recheck and update attribution |
| Number or requirement | Are units, method, location, and applicable edition clear? | Verify or remove |
| Illustration | Does the caption support the actual claim? | Relabel or replace |
| Reader action | Could the wording be mistaken for a universal design instruction? | Add conditions and decision boundaries |
For consequential statements, check the primary basis where possible: inspection records for a described case, the underlying study for a research claim, or the applicable current document for a referenced requirement. A newer source does not necessarily supersede an older one. Compare test conditions, materials, loading, and outcome measures. If the original evidence cannot be recovered, support the claim independently or delete it. Years on the site do not make an unsupported assertion reliable.
Preserve the explanation readers still need
The most valuable part of an older technical article is often its causal reasoning. An account of pavement distress, for example, might connect visible deformation to traffic loading, temperature, water, and support conditions. Do not replace that chain with a catalogue of new materials. Check instead where it needs qualification: whether several causes could produce the same distress, what field evidence might distinguish them, and where laboratory results may not predict in-service behavior.
Before rewriting, list what should stay: accurate definitions, diagrams that still explain a mechanism, examples with documented conditions, and cautions against overgeneralization. Then list what must change. A shorter, smoother draft is not an improvement if it strips out the detail that made the original useful.
Protect context around examples
A case study is not a design rule. If an embankment example reports settlement after construction, keep the ground conditions, construction sequence, observation period, and measurement limits needed to interpret it. Do not lift out a single settlement figure and present it as typical performance. If those details are unavailable, describe the example cautiously or omit the number.
Keep uncertainty visible
Newer evidence may narrow a conclusion rather than strengthen it. If monitoring data cover only a short period, say so. If a technique performed well under particular exposure conditions, distinguish those conditions from others. Do not change “may indicate” to “indicates” just to make a sentence sound firmer. Precision includes stating what the evidence cannot establish.
Update for the task, not for a fashionable term
Readers tend to arrive with a practical question: what a symptom might mean, what evidence to collect, or how to interpret a method’s limits. Check that the article answers its question near the top. A page about bridge inspection methods should not bury the difference between detecting an anomaly and confirming its cause under a long account of new instruments.
Organize the revision around the reader’s decision. For drainage-related pavement damage, that could mean observed symptoms, plausible water pathways, evidence needed to distinguish them, and limits of each intervention. A paragraph on recent monitoring technology belongs only if it explains how diagnosis changes. If the original scope still works, revise the relevant sections rather than changing the article’s central promise.
Give related pages clear boundaries, too. A materials article need not become a second inspection guide merely because inspection findings inform material selection. Where a link helps, use it to direct readers to the adjacent question—for example, [pavement material selection by failure mechanism]—rather than repeating that analysis. Remove the square brackets during editorial formatting if the publication’s link style does not use them.
Make substantive changes visible
A silent rewrite can confuse readers who saved or quoted an earlier version. Keep the original publication date separate from the revision date. Add a brief update note when a change affects interpretation: a corrected number, a revised account of a method’s limits, or the removal of advice that no longer holds. Spelling fixes do not need a prominent notice.
Say what changed instead of announcing that the page was “refreshed.” For example: “Updated to distinguish screening measurements from confirmatory investigation and to remove an unsupported performance comparison.” If an earlier error could have influenced decisions, state the correction plainly. Keep an internal record of the old wording, the reason for the change, the evidence checked, and the reviewer. It will help editors answer questions about older quotations.
Do not change a URL merely to reflect a new year or title. Bookmarks, references, and internal navigation may depend on it. If a fundamental change in scope requires a move, plan the redirect and inspect incoming links and navigation text. A redirect will not fix an anchor that promises a subject the revised page no longer covers.
Give technical review a defined scope
An editor can improve clarity, but changes to load paths, geotechnical assumptions, diagnostic thresholds, material performance comparisons, safety implications, or compliance claims need subject-matter review. Ask the reviewer to consider the surrounding context as well as the revised passage. A technically correct sentence can mislead if the paragraph before it implies that it applies more broadly.
Make the request specific. Provide the old and proposed wording, the basis for the change, and any uncertainty. Ask, for example, “Does this observation distinguish drainage failure from subgrade weakening?” or “Does the cited test support the claimed field outcome?” “Check accuracy” gives a reviewer much less help in finding the highest-risk problems.
Follow technical review with a separate editorial pass. Check that headings fit their sections, units are consistent, figures are legible, and terms have not shifted meaning. Read lists of recommended actions out of context; add conditions wherever an isolated excerpt could be mistaken for universal project guidance.
Check the published page, not just the draft
Publication can introduce errors of its own. Tables may lose column headings on narrow screens, captions may separate from figures, and an update note may appear below the outdated text it qualifies. Preview the finished page on desktop and mobile. Check image alt text, internal anchors, downloadable material, and excerpts used elsewhere on the site. If the title or opening changed, those excerpts must still describe the page accurately.
Set the next review according to the article’s risk. Revisit a durable explanation of a mechanism when evidence challenges it or a reader flags an ambiguity. Review pages about product capabilities, active requirements, or operational procedures more deliberately. For the next page in the queue, write one sentence stating its enduring claim and another naming the claim most likely to age. Those two lines give its next audit a useful starting point.
