Why we tested this scenario
In subsea pipeline operations, one of the main challenges is not the lack of understanding of risks, but the limited observability of certain types of events.
There is a category of external impacts that:
- occur over a short time,
- do not cause immediate loss of containment,
- do not trigger standard alarm systems,
- but introduce delayed risk in the form of coating damage, local deformation, or accelerated corrosion.
These events are particularly problematic because they often go unnoticed at the time they occur and only become evident much later, when the link to the original cause is difficult to establish.
As part of validating InfraScan’s monitoring approach, we wanted to answer a practical engineering question:
Can Distributed Acoustic Sensing detect and logically reconstruct an external mechanical scenario using only physical acoustic data – without AIS, video, or contextual assumptions?
Below we describe how a typical external interference scenario manifests itself in DAS data and how it can be interpreted from an engineering perspective within a controlled test framework.
Overall appearance of the test signal
At a high level, the DAS data shows a clear and consistent sequence:
source motion → abrupt change of regime → localized interaction → renewed motion with interruptions
We do not claim that this represents a real incident. The purpose of this test is to demonstrate how such a scenario would appear in DAS data if it occurred.
Phase 1. Source movement along the pipeline corridor
At the beginning of the analyzed time window, the system detects a series of events classified as moving.
On the time – distance plot, this appears as:
- well-defined diagonal signatures,
- stable direction of propagation,
- high repeatability across adjacent segments.
The estimated source velocity along the corridor is on the order of tens to over one hundred meters per second, which is consistent with a large moving source influencing the pipeline environment.
From an engineering standpoint, this represents a normal background condition associated with surface activity above the pipeline. At this stage, there is no immediate threat to pipeline integrity.
Phase 2. Abrupt signal change – transition of interaction regime
Further into the time window, a distinct change in signal behavior becomes apparent.
The diagonal signature:
- changes slope abruptly,
- becomes partially discontinuous,
- and acoustic energy starts to localize instead of propagating smoothly along the corridor.
In the automated event log, this appears as a transition from moving events to short-duration, quasi-stationary detections.
Viewed from an engineering perspective rather than an algorithmic one, this suggests a simple explanation:
the source has significantly slowed down or stopped and has begun interacting directly with the surrounding environment.
Within the scope of this test, this pattern represents a transition into a localized mechanical interaction regime.
Phase 3. Localized interaction – why the algorithm labels it as “leak”
Following the regime change, DAS detects a series of localized events that the classifier labels as leak.
It is important to clarify upfront: within this test, these events are not interpreted as an actual leak.
The classification outcome is understandable from an algorithmic perspective:
- the source is nearly stationary,
- the signal is persistent,
- the frequency content is broadband,
- and the energy is spatially confined.
For a classifier, this combination resembles a leak-like acoustic signature. For an engineer, however, it is a typical acoustic fingerprint of mechanical contact.
In the context of this test, these signals are interpreted as a model of:
- friction,
- intermittent impacts,
- localized vibration of the pipeline,
i.e., external mechanical contact with the pipeline.
This is the point at which the first meaningful integrity risk appears, even though there are still no indications of loss of containment.
Phase 4. Renewed motion with interruptions – the most sensitive regime
After the localized interaction phase, the signal evolves again.
The system detects moving events once more, but:
- the motion is less stable,
- energy appears in intermittent bursts,
- activity remains concentrated near the previous interaction zone.
From an engineering standpoint, this pattern is consistent with a scenario in which:
- motion resumes,
- the external interaction is not fully disengaged,
- repeated short-duration contacts occur along the pipeline.
This regime is often considered more critical than a single impact, not because of peak loads, but due to the potential for:
- coating damage,
- local plastic deformation,
- accumulation of delayed defects that may manifest later in the pipeline’s life.
Why the test risk level is MED, not HIGH
It is important to be transparent about the limits of the technology.
DAS:
- does not measure residual deformation,
- does not assess wall thickness,
- does not observe internal pressure or flow conditions.
For this reason, the system does not declare an аварийный or critical state. Instead, it does exactly what it is designed to do:
identify a potentially hazardous external process and provide a basis for engineering follow-up.
What this test demonstrates
This test does not claim that an incident occurred. It demonstrates that InfraScan is capable of:
- detecting dynamic external activity,
- identifying transitions in interaction regime,
- recognizing localized external interference,
- reconstructing a coherent sequence of events,
- doing so before any loss of containment and without human intervention.
This capability is the core practical value of DAS-based monitoring.
Short conclusion
infrascan.ai does not attempt to predict failures. It measures physical processes.
This test scenario illustrates that even without auxiliary information sources, DAS can provide engineers with sufficient insight to identify potentially harmful external interactions early and to act before those processes evolve into actual failures.
