FAT/SAT roles, scripts, and evidence capture.

Independent eyes validate the system. Define scope, roles, and prerequisites, then build scripts/forms that mirror your interlock matrix (352), I/O list (523), and alarm philosophy (536). Evidence capture standards (716) and clear sign-off criteria prevent disputes. Manage NCRs (719) and punch lists to closure, and assemble final report packs that strengthen HVM bollard approvals and crash rated bollard handovers (736, 638). For context and deeper integration topics, see this section and the chapter hub. When UAE approvals apply, consult SIRA Bollards (UAE) for jurisdictional expectations.

Important: This is a general guide. For live projects we develop a tailored Method Statement & Risk Assessment (MS/RA) and align with authority approvals (e.g., SIRA) where in scope.

538.1 Scope & responsibilities

Define what’s witnessed and by whom. Clear scope aligns HVM bollard stakeholders (631–637).

Start by declaring which subsystems are in scope for SAT versus FAT: bollard drives, control panels, safety devices & measures, fire/BMS/SCADA interfaces, and access/CCTV coordination. Clarify boundaries: what counts as a “rating-critical dependency” (e.g., footing spec, rebar, grout) and what’s operational. Assign accountable owners using a simple ITP with witness points.

Publish a one-page “who does what” list covering the client, contractor, integrator, and third-party witness. Reference pre-commission pages (631–636) so everyone brings the right tools and data. This avoids scope creep during the session and keeps the third-party focused on verification rather than discovery.

AspectWhat mattersWhere to verify
PerformanceTested system (bollard + footing)Global crash ratings
OperationsDuty cycles, fail-state, safetyInstallation (hub)

538.2 Pre-requisites

Close defects, gather evidence, and freeze versions (716, 537). Prereqs prevent crash rated bollard surprises.

Before the witness arrives, run a Pre-Commission Checklist and ensure a formal design freeze. Collect as-built drawings, cable schedules, loop test results, panel photos, and calibration sheets per Evidence Capture Standards. Confirm the I/O list and interlock matrix match the installed configuration.

538.3 Test scripts & forms

Mirror interlock matrix and modes (352, 525). Scripts make HVM bollard tests reproducible (634).

Author step-by-step scripts that reference specific tags and alarm IDs, including normal, maintenance/bypass, and emergency/EFO modes. Each step should state initial conditions, the action, the expected indication (HMI lamps, messages), and the acceptance band for timing or travel. Link each row back to the matrix and tag names for traceability.

Use standardized forms (tick boxes + comments) so multiple lanes and arrays can be tested consistently. Plan a short incident response drill to prove operator recovery hints and reset hierarchy after a latched fault.

538.4 Evidence capture

Standard photo/video and logs (444). Evidence convinces crash rated bollard reviewers.

Define how to capture photos (wide→detail sets), screen recordings, and raw logs. Use time-sync (NTP) so camera timestamps and PLC/HMI logs align. Store files against the script step IDs and maintain an evidence chain with signatures. For UAE work, keep a SIRA-friendly folder structure and ensure personal-data handling respects site policy; if in scope, note it in the submission cover.

538.5 Witness roles

Assign lead, recorder, and approvers. Roles keep HVM bollard sessions orderly.

Nominate a session lead (conducts), a recorder (completes forms, timestamps media), and approvers (client/Owner’s Engineer). The third-party witness confirms conformity, not design intent. Publish a schedule, book utilities access, and confirm witness forms are pre-filled with project metadata to avoid clerical errors.

538.6 Nonconformance handling

NCR, root cause, fix, retest (719). Handling maintains crash rated bollard confidence.

Classify findings immediately: observation, minor, or NCR. Log them in the NCR register with root cause, corrective action, and retest step. If a safety inhibit fails, mark the session as conditional and plan a targeted re-witness on only the affected steps, keeping the evidence pack whole.

538.7 Sign-off criteria

Numeric thresholds and acceptance bands (714). Criteria make HVM bollard pass/fail clear.

Write pass/fail rules that use measurable bands: rise/fall time, lamp/annunciation latency, loop detect stability, KPI thresholds. Avoid ambiguous phrases like “works normally.” If blue-light overrides or egress priorities apply, show precedence with a simple priority matrix and include timing bands for conflicts.

538.8 Punch list tracking

Log, assign, and close actions. Tracking shortens crash rated bollard closeout (639).

Keep a single punch list (snag list) with due dates and owners. Group items by lane/panel to plan efficient return visits. After fixes, re-run only the affected script steps and append new evidence without overwriting the original session. Closeout rolls into the Snagging & Handover process and final retention.

538.9 Final report pack

Bundle scripts, results, evidence, approvals (736). Pack completes HVM bollard acceptance (938).

Deliver a numbered PDF set: signed scripts, results summary, NCR log, evidence index with links, and the signed acceptance sheet. Cross-reference the Handover Pack Index and include a “reader guide” page so reviewers can quickly locate the step, lane, and media. For authority submissions, follow the Submission-Pack Guidance formatting and indexing rules.

Related

External resources

538 Third-Party Acceptance Tests — FAQ

What’s the difference between FAT and SAT for HVM bollards?
FAT verifies panels, software, and wiring looms at the factory; SAT verifies the fully installed system on site, including interfaces, safety devices & measures, and operator workflows. SAT evidence becomes part of the handover pack.
Who should attend a third-party witness session?
Typically: the third-party witness, the contractor/integrator test lead, the recorder, and client/Owner’s Engineer. Specialists (fire, BMS, access control) join for their scoped tests only.
What if a test step fails during witnessing?
Record the failure as an observation or NCR, capture evidence, and agree corrective actions with a dated retest. You can re-witness only the affected script steps once the fix is proven.
What goes into the final acceptance report?
Signed scripts, results summary, NCR/punch list closeout, photo/video/log index, and the signed acceptance sheet—organized per the Submission-Pack Guidance for easy review.