2025–2026
Learning signalling properly, then automating its verification
Control table testing and principle tester rules first; automation across mainline and yard second — in that order, deliberately.
- CBTC
- Control table testing
- Principle tester rules
- Verification automation
- Python
The problem
CBTC verification is repetitive at volume and unforgiving of inconsistency. Testing effort across mainline and yard consumed specialist time that was in short supply, and how a test was executed and recorded varied more between people than it should.
What I did
I did not start by writing code. I trained under the programme’s Signalling Director in principle tester rules, reading signalling plans, control table testing and CBTC train control until I could do the verification by hand. Only then did I build automation to reduce the manual effort and standardise how tests are executed and evidenced.
What changed
Reduced manual verification effort and improved consistency of testing across mainline and yard, and extended my own scope on the programme from commissioning into signalling oversight.
What I took from it
I do not automate what I do not understand. Learning to test a control table by hand before building anything is the reason the tooling was trusted by the people who had to rely on it — and it is the approach I would take into any unfamiliar system.