REEKON
Fire Data Analyst — Pattern Intelligence
Watch REEKON CORE ingest 90 days of RMS incident data, identify a statistically significant arson cluster, cross-reference with historical patterns and geographic data, and generate an actionable brief for the arson investigation unit — automatically.
8 steps · ~40 seconds · RMS arson pattern analysis
REEKON
Fire Data Analyst
Ready
—
← Back
RMS pattern analysis engine
Pattern confidence
—
RMS data CAD history GIS / cluster Historical match Statistical AI pattern
📊
Step 1 — RMS data ingested
REEKON ingests 847 incident records from the agency RMS system covering the prior 90-day window: incident type, location, time, cause determination, and unit response data. Initial anomaly scan initiated.
847 records ingested90-day window
🚪
Step 2 — CAD cross-reference
CAD historical records cross-referenced with RMS data. Response times, unit assignments, and dispatch narratives added to incident dataset. 14 undetermined-cause fires flagged for deeper analysis.
14 undetermined-cause flaggedCAD matched
🔎
Step 3 — Geographic cluster detected
GIS spatial analysis identifies statistically significant cluster: 14 of the flagged incidents fall within a 0.8 sq mi zone in the northeast district. Cluster density: 6.2× above baseline for comparable zones.
Cluster detected — 0.8 sq mi6.2× above baseline
🕑
Step 4 — Temporal pattern identified
Time-of-day analysis reveals 11 of 14 cluster incidents occurred Tuesday–Thursday, 0200–0400 hrs. Day-of-week and time distribution is statistically non-random (p < 0.003). Probable intentional pattern.
Tue–Thu 0200–0400p < 0.003 — significant
🔍
Step 5 — Historical MO match
REEKON cross-references against historical arson series database. Finds 88% MO similarity to the 2019 Eastside series (eventually solved — 4 incidents, same time pattern, same zone type). Pattern scoring elevated.
88% match — 2019 Eastside seriesHistorical DB match
🌟
Step 6 — Statistical significance confirmed
Multi-factor pattern analysis complete. Cluster confidence: 97.3% — exceeds threshold for arson investigation referral. REEKON drafts incident brief with cluster map, timeline, MO comparison, and recommended patrol resource request.
97.3% confidenceReferral threshold exceeded
📄
Step 7 — Actionable brief generated
REEKON generates full analysis brief: cluster map overlay, incident timeline, MO scoring comparison, patrol resource recommendation, and community notification draft. All data linked back to source records in RMS and CAD.
Brief generatedSource-linked throughout
📲
Step 8 — Arson unit notified
Pattern brief routed to Arson Investigation Unit, Fire Marshal, and Data Analysis Lead. Includes direct links to all 14 source incident records, GIS cluster export, and patrol resource request draft — ready for supervisor signature.
Arson unit notified14 records linked
0–15 min (core)
15–25 min
25–30 min
Perimeter
Northeast District — Los Angeles County
REEKON CORE — Alert detail
📊
REEKON CORE
Data analyst brief
Awaiting pattern analysis completion...
📊
REEKON CORE — Alert detail
AI pattern analysis
Populates on pattern threshold...
Generating REEKON assessment...
Pattern summary ✓ Statistical significance
Incidents analyzed847 (90-day window)
Arson cluster identified14 incidents
Geographic radius0.8 sq mi
Time patternTue–Thu 0200–0400
Confidence (statistical)97.3%
Comparative analysis
vs. prior 90-day period+340% incident rate
vs. county average6.2× above baseline
Similar pattern match2019 Eastside series
MO similarity score88%
Recommended actions
Share with arson investigation⇨ Queued
Patrol resource requestDraft ready
Community notificationPending approval
Authoritative sources
📊 RMS — 847 incident recordsView →
🚪 CAD historical — same periodView →
🔍 GIS cluster analysisView →