Strengthening ATM Resilience through Smarter Monitoring
At ATSEP Seminar 2025, Callum Winship shared how smarter monitoring can help
ATM move from reacting to alarms to anticipating issues.
The idea began with a quiet Saturday
that turned hectic in seconds, a surge of alerts and a team working fast to stabilise systems. The question that followed was simple and powerful.
Can we see weak signals early enough to act before a fault unfolds?
Callum’s team teaches systems what normal looks like, then watches for small departures from that baseline. In one case the method flagged unusual log activity five days before a failure in a key system. In another it surfaced issues that traditional tools had not captured. The aim is not to replace human judgement but to give teams time and context, so action can be taken while margin still exists.
Trust sits at the centre of this work. Any solution must be explainable, testable and safe.
That is why the first applications will be in validation environments, where updates and configurations can be exercised before release. Over time,
when evidence is strong and procedures are clear, this kind of monitoring can add a further layer of protection in operations.
For ATSEPs the craft evolves rather than changes identity. Less energy spent chasing alarms. More focus on interpreting patterns, spotting drift and choosing the right interventions.
As Callum said, “Technology will not take anything away from what ATSEPs do today, it will simply make what they do even safer.”
The direction is clear. Early pattern detection buys time and reduces risk. Better data capture and smarter analysis make monitoring more useful. Keeping people at the centre ensures accountability and sound decisions. Together, these elements point to a more predictive and resilient approach to safety in ATM, where human expertise and technology support each other.
If you would like to know more about the ATSEP seminar, subscribe to our newsletter to get the series, or contact us if you want to explore how this approach could fit your context.