CBTC Testing & Commissioning Checklist
Testing and commissioning is the most consistently underestimated phase of a CBTC deployment. A project can look complete on paper — equipment installed, software loaded, interfaces connected — and be nowhere near ready for revenue service. This sheet is the phase structure, the durations that hold up in practice, and the items each phase has to close before the next one is allowed to start.
Free to use and share with attribution. No email required.
What's inside
- All six phases with purpose, performer, environment and realistic duration in one table.
- Per-phase sign-off checklists you can print and tick — FAT through revenue service demonstration.
- Shadow-running expectations: why 6–12 months, and what nightly log analysis is looking for.
- The mixed-mode capacity trap: why the least-capable train type governs headway.
- The FTA System Integration Review gate and the evidence a PMOC actually asks for.
The durations are the useful part
Most published CBTC test plans are structurally right and temporally fantasy. The numbers on this sheet come from documented US practice: 6–12 weeks of factory acceptance testing; 12–20 weekends of site integration testing, because that is the access a live railway grants — NYCT could reach the track from Friday night to Monday morning, about five to seven uninterrupted hours at a time; 4–8 weeks of dynamic testing; and 6–12 months of shadow running. The L Line shadow-ran roughly eight months before cutover.
Simulation moves those numbers but does not remove them. The 7 Line project ran a full-system simulator with production software on realistic track topology, validating thousands of scenarios before a test train moved in Manhattan. That cuts on-track testing from roughly 12–16 weeks to 6–8. It cannot replicate radio propagation in an urban canyon, environmental noise, or how an operator actually behaves at 06:40 — which is why the remaining weeks are the essential ones.
One trap worth naming
In mixed mode, the least-capable train type governs the headway. If legacy trains hold three minutes, CBTC trains hold three minutes too. A line engineered for 30 tph under full CBTC will not exceed roughly 20 tph while mixed running lasts — so the mixed-mode period has to be planned and budgeted as a degraded-capacity period, not presented to stakeholders as an early delivery of the benefit.
Sources & method
Phase structure and durations from Communications-Based Train Control, Volume 1, Chapter 12.5 (Table 12.6), by Chunjun (Francisco) Wang. Durations reflect US brownfield practice on mid-size lines; greenfield projects compress SIT and parallel operations substantially.