Context
Challenge 1 explicitly lists disqualified methodologies (real-time vessel monitoring, anomaly detection on AIS/satellite fusion, vessel behavior profiling, risk scoring, geofencing, conventional sanctions screening, off-the-shelf behavioral analytics). The submission uses some of these terms in the narrative, which could cause evaluators to flag it as borderline non-compliant.
Current state
A defensive clarification appears late in the narrative. No structured table or early section directly maps arktrace's approach against each excluded methodology.
Required
Add a dedicated subsection — "Alignment with Challenge Exclusions" — early in Annex A (after the Relevance section) with a table mapping each disqualified methodology to how arktrace differs:
| Disqualified methodology |
How arktrace differs |
| Relying solely on historical data |
Continuous 15-min re-scoring on live AIS; 60–90 day pre-designation lead from causal model |
| Real-time vessel monitoring |
Not a tracker — a prediction engine ranking candidates by causal evasion likelihood |
| Anomaly detection on AIS + satellite |
AIS/EO feed the causal model as evidentiary substrate; primary discriminator is DiD ATT |
| Vessel behavior profiling |
HDBSCAN used only to define normal MPOL baseline; discriminator is causal ATT, not profile distance |
| Risk scoring |
Composite score weights derived from ATT significance testing, not expert heuristics |
| Geofencing |
Not used as primary signal; geographic scope is a pre-defined bounding box only |
| Conventional sanctions screening |
Sanctions list is ground-truth for DiD event timing, not the detection mechanism |
| Off-the-shelf behavioral analytics |
Entire pipeline built in-house; no commercial platform provides DiD causal inference |
Acceptance criteria
Context
Challenge 1 explicitly lists disqualified methodologies (real-time vessel monitoring, anomaly detection on AIS/satellite fusion, vessel behavior profiling, risk scoring, geofencing, conventional sanctions screening, off-the-shelf behavioral analytics). The submission uses some of these terms in the narrative, which could cause evaluators to flag it as borderline non-compliant.
Current state
A defensive clarification appears late in the narrative. No structured table or early section directly maps arktrace's approach against each excluded methodology.
Required
Add a dedicated subsection — "Alignment with Challenge Exclusions" — early in Annex A (after the Relevance section) with a table mapping each disqualified methodology to how arktrace differs:
Acceptance criteria