WELCOME to the New In Signal
by author Ivan Ristić...
Promo & Careers | Contact us directly for more info how to advertise for FREE!
Latest News
TRACER: STREAMLINING TEST RECORDS FOR SAFETY-CRITICAL COMMISSIONING | Author: James Sefton, IRSE Licensed Signalling Engineer, Director Sefton Rail Engineering Ltd / Founder Kontinuity
I was brought in to test a system that should have been ready. It wasn’t.
Multiple wiring errors. Installation faults that should have been caught before handover to test and commissioning. I knew the functional behavior of the system, what it was supposed to do, but I didn’t know the schematic details.
So I fault found it. On a program with tight deadlines and test tracks booked, on equipment I hadn’t been briefed on at the drawing level, under pressure to report daily progress to a project team that wanted percentages. This is a good way to learn a new system, but not a good way to fix things fast.
This got me thinking about maintenance tasks where faults and trains out of action can have serious cost implications. Being unable to provide full service can hit train and freight operating companies hard.
From this, Tracer was born and has now evolved from a maintenance tool to a maintenance and commissioning tool for safety-critical systems.
The record problem
Railway testing and commissioning is meticulous work. Every test has a procedure. Every result must be recorded. Every failure must be logged, investigated, resolved, and documented before the system is signed off. The audit trail is what proves the work was done correctly, not the engineer’s recollection.
In practice, that audit trail is often a ring binder, a spreadsheet updated at the end of the day, and whatever the engineer on shift can remember clearly enough to write down after ten hours on the tools.
“End of a commissioning shift: there are test certificates to fill in, test logs to record issues found, a site report to compile, and a project manager asking what percentage complete the program is. All of it was done manually. All of it was done from memory or notes taken during the day. All of it carries the human error that comes in when the person doing it is tired, and the information is already hours old.”
This is not a criticism of the engineers doing it (I am one of them). It is a description of the tooling available to them. Nobody has built something better, because the people who understand the problem well enough to build the right tool are the ones on the vehicles, not writing software.
I’ve spent many years in railway signalling – ETCS, CBTC, relay-based systems across major programs in London and elsewhere. I’ve written those end-of-shift reports. I’ve been in the update meetings to pass progress up the chain. I’ve been the engineer redoing testing or fault finding that’s already been completed, but was missed at handover.
At some point, I stopped accepting that this was just how it worked.
What Tracer does
Tracer is a real-time test record tool for safety-critical commissioning. It has two modules. Tracer Maintain is a fault-finding tool built to assist maintenance technicians and record fault-finding outcomes and progress. Tracer Test digitalizes test records and processes, and reports live progress to project teams.
Fault diagnosis module
Guided fault finding for relay circuits. The engineer loads the relevant circuit – ETCS, CBTC, relay-based, and works through a step-by-step procedure on their phone or tablet. Each step shows the instruction, the measurement point, the expected result, and highlights the relevant part of the circuit diagram. Tap any component to start fault-finding from that point. Navigate bidirectionally between relays and contacts without losing your place.
Tracer Maintain: guided fault diagnosis for relay circuits. Each step highlights the relevant circuit element and specifies the measurement expected.
Every session is recorded: engineer ID, circuit, steps taken, outcome, conclusion, and severity. When the fault find is complete, the record exists — timestamped, attributed, stored.
The module is mobile-first and designed for use on the vehicle. PIN authentication and competency confirmation. It runs as a Progressive Web App – load it once, and it works regardless of what the network is doing. Rail environments don’t offer guaranteed connectivity. At depots, tunnels, remote sites, vehicles mid-test, the mobile signal is unreliable or absent. A tool that depends on a live connection fails at exactly the wrong moment. Tracer doesn’t. The session records are there when the program manager asks what happened on that fault six months later, and can be uploaded onto the asset management tool (Maximo or equivalent).
Commissioning module
Structured test execution with a workflow enforced by the system, not by policy.
Sessions move through defined states: a tester executes steps, an analyst reviews, and a Tester in Charge signs off. The system prevents an engineer from analyzing or signing off on a session they contributed test steps to – role conflict enforcement that is not a reminder or a guideline, but a system-level constraint.
Tracer Test: session complete view. Failed and blocked steps are recorded with mandatory reasons. The TIC cannot sign off until all steps are accounted for.
Blocked steps are recorded with a mandatory reason, a timestamp, and a resolution note. Duration calculated automatically. The block record exists whether or not anyone writes it down.
When the TIC signs off, they use a PIN. The timestamp is written at that moment. It cannot be backdated.
Certificate generation, site reports, and test logs are produced automatically from the session record. Not compiled from notes at the end of a shift. The record is built as the work is done.
The tracker problem
The other part of the problem is visibility and the overhead it creates. A TIC running a commissioning program typically maintains a detailed tracker alongside everything else: what’s been tested, what’s blocked, what’s failed, what’s outstanding. That tracker is built from manual data input, session by session, shift by shift. It takes time. It duplicates work that has already happened during testing. And it still carries human error, because it is only as accurate as the last person who updated it. Mistakes are not difficult to make, I’m speaking from experience.
Tracer Test: session overview showing status, contributing testers, sign-off timestamp, and automatic document generation – certificate, session report, and site report.
Tracer eliminates that entirely. The commissioning dashboard is built from the session records in real time. Completed steps, blocked tests, failures, and durations are all captured automatically as the work is done. The TIC doesn’t maintain a tracker because the tracker maintains itself. The PM doesn’t call for a progress update because the progress is already visible. The record is the work, not a summary of it written afterward.
Field Pack
For air-gapped environments – vehicles without network access, a Field Pack exports a self-contained offline file. The engineer works through the session, the pack records everything locally, and when they’re back in range, it merges into the main record. Three execution modes: fully online, fully offline, and split.
Where it is now
The fault diagnosis module has an MVP. The commissioning module is in active development.
Neither has been tested on a live project yet. I am still doing manual reporting on the jobs I’m working on while I build this. It is also the reason the design decisions in Tracer reflect what the work actually involves – the role conflict enforcement, the state machine that cannot be skipped, the block tracking with mandatory duration, because they came from being in the environment.
Where it’s going
As I’ve built Tracer, the opportunities for further integration become clearer. The next direction worth exploring: coupling Tracer to a hardware interface that connects directly to a passenger or freight vehicle to automate functional wiring verification before a subsystem connects. Prove the wiring works functionally, before the signalling system sees it.
This doesn’t replace verification testing. Functionally correct wiring is not the same as wiring that matches the drawing. But it changes what arrives at T&C. Fewer wiring faults are absorbed by the commissioning program. Cleaner handover. A record of what was checked before the T&C team took over.
James Sefton is an IRSE-licensed signalling engineer. He is the founder of Kontinuity and the developer of Tracer. He is contactable via LinkedIn (linkedin.com/in/james-sefton-primaresi) or at james@kontinuity.co.uk
Subscription Plans
Free
Subscribe Now for FREE
- Free registration
- Timely updates
- Free Handbook Download
- Access to Newsletter Issues
- Access to introduction sections
Basic
Monthly fee 9,99 CHF
- Free registration
- Timely updates
- Free Handbook Download
- Access to Newsletter Issues
- Free access to the paid articles
Premium
Monthly fee 12,99 CHF
- Free registration
- Timely updates
- Free Handbook Download
- Access to Newsletter Issues
- Access to introduction sections
- Free access to the live events with our top authors
