Skip to content

Safety and Accessibility Review

Assessment review — instantiates Revealed-Use Path Alignment

Tests a candidate accommodation against the non-negotiables — safety, accessibility, dignity, privacy, environment — before anyone decides to pave, reroute, or block the revealed path.

A trace tells you what people do; it does not tell you what you may honor. Safety and Accessibility Review is the gate between the two. It takes a candidate accommodation — "make this shortcut official" — and runs it through a battery of non-negotiable tests: is it safe, is it accessible, does it preserve dignity, privacy, and environmental constraints? The output is a reasoned verdict — legitimize, reroute, or block — with the reasons attached. Its distinguishing role is to be the archetype's guard against naïve "pave the cowpaths" thinking: it is the difference between paving a desire path and paving one that happens to run across a live rail line. It judges legitimacy; it does not capture the trace and does not enact the outcome.

Example

At a regional rail station, commuters routinely cut across the tracks at the far end of the platform rather than walk to the footbridge — a heavy, unmistakable desire path. The review does not ask whether the shortcut is popular; that is not in question. It asks whether it can be made legitimate. Safety: crossing live track — an unqualified no. Accessibility: the footbridge is stairs-only, which fails wheelchair users, travelers with luggage, and parents with strollers — and that failure is very likely why some people risk the tracks in the first place. Precedent: legitimizing an at-grade crossing here invites the same everywhere on the line. The verdict comes back block the crossing — but the review's real yield is the finding underneath it: the genuine, legitimate need is for a level, accessible crossing, and that need must now be served some other way. Verdict plus reasons hand off to whoever enacts the change.

How it works

What distinguishes this review from a capture step or a decision step is that it applies fixed tests and returns a reasoned verdict on a specific candidate:

  • A battery of non-negotiables. Safety, accessibility (anchored to a real external standard), privacy, environment, and equity are each run against the concrete candidate accommodation.
  • Any test can veto. A single failed non-negotiable can block, so the review is not a popularity vote or a cost-benefit tally.
  • Reasons, not just a ruling. The output carries why, so a "block" still surfaces the legitimate need underneath — the input a good alternative is built from.

It is also where the people absent from the trace — the ones the current design already excludes or endangers — get counted.

Tuning parameters

  • Test battery — which non-negotiables are in scope and in what order. A missing test (environmental impact, say) is exactly where a harm slips through.
  • Veto versus weigh — whether one failing test blocks outright or trades off against others. Hard vetoes protect safety but can over-block benign shortcuts.
  • Evidence bar — how much proof each test demands before it passes. A high bar is rigorous but slow; a low bar rubber-stamps.
  • Whose-safety frame — whether the review counts only current users, or also bystanders and the deterred. The narrow frame is the classic blind spot.
  • Anchoring standard — which external standard defines "safe" and "accessible." A weak or in-house standard quietly lowers the bar to whatever is convenient.

When it helps, and when it misleads

Its strength is that it is the discipline standing between a real usage signal and a reckless response: it stops the organization from legitimizing a shortcut that is dangerous, exclusionary, or unlawful, and it forces the absent user — the one who left no trace because the design already shut them out — into the decision.

Its failure modes are the mirror image of that strength. Reduced to a formality, it becomes a rubber stamp blessing an accommodation already decided on other grounds. Run too conservatively, it over-blocks — treating every deviation as a hazard and killing the benign, useful shortcuts the archetype exists to learn from. And framed too narrowly, it privileges the visible current users over the absent ones. The discipline is to anchor each test to a real external standard rather than in-house comfort,[1] require written reasons for every verdict, and ask explicitly who is missing from the trace.

How it implements the components

Safety and Accessibility Review realizes the filter side of the archetype's machinery — the legitimacy tests, not the capture or the change:

  • safety_and_legitimacy_filter — its core: the battery of safety, compliance, and environmental tests that yields the legitimize / reroute / block verdict.
  • equity_and_visibility_check — it weighs who the path serves and who it excludes or endangers, deliberately counting users absent from the trace.
  • privacy_and_consent_guardrail — it tests whether honoring — or instrumenting — the path would breach privacy or consent.

It does not capture the trace it judges (Use-Trace Heatmap and Workaround Inventory supply that) and it does not enact the verdict — blocking-with-a-substitute is Route Closure with Alternative; legitimizing the path is Informal Route Legalization Patch.

Notes

This is a trace-specific review, not a blanket safety audit: it judges one revealed path against the non-negotiables, not the whole system in the abstract. That focus is what keeps it inside the archetype rather than drifting into general compliance sign-off — and it is why a "block" verdict is expected to name the legitimate need underneath, so a downstream mechanism can serve it rather than simply erase it.

References

[1] Universal design — the principle, associated with architect Ronald Mace, of designing environments to be usable by the widest range of people without special adaptation. Anchoring the accessibility test to it (rather than to a minimal in-house checklist) is what stops the review from certifying a route that merely works for the median user. ↩