CBRN-CADS v3 — The 11-Step Decision Tree from Detection to After-Action Report

The hardest problem in chemical and biological response is not detection. It is the structured handoff between civilian first responders, military CBRN specialists, environmental authorities, and intelligence services during the first 90 minutes of an event. The National Disaster Management Institute's 2013 "Chemical Weapon Attack Action Sequence" (NDMI Figure 3-42) defined an 11-step civil-military response sequence that has remained the Korean reference architecture for over a decade.

CBRN-CADS v3 operationalizes that 11-step sequence into an AI-augmented decision support system. Each step is a decision node with explicit inputs, AI-recommended outputs, and audit-grade handoff documentation.

The 11 Steps, AI-Augmented

| # | Step | AI Decision Node | Live Inputs | |---|---|---|---| | 1 | Situation Onset | Sensor stream ingestion + automatic agent identification | Multi-modal: video, gas chromatograph, IMS, atmospheric | | 2 | Alert Propagation | 112/119 auto-bridge + village PA + mobile push | Geographic radius, initial agent estimate | | 3 | Evacuation / Protection | Mask donning prompts + evacuation route generation | Wind direction, terrain, population density | | 4 | Casualty Triage | MARK-1 auto-recommendation + hospital matching | Contamination concentration, casualty count | | 5 | CBRN Control Center | Four-agency coordination protocol initiation | Roles for NIS, military, police, environment | | 6 | Agent Detection | Sensor grid auto-deployment | Wind cross-section sampling priorities | | 7 | Regional Control | Police line + population displacement protocols | Real-time plume model | | 8 | Decontamination | STB/DS2 reagent auto-selection | Agent class + surface type + weather | | 9 | Verification | Sensor residual concentration validation | Atmospheric clearance time estimation | | 10 | After-Action | Casualty management + logistics support | Patient tracking + supply chain | | 11 | Feedback / Learning | AI model update | Scenario library extension |

Three External System Handoffs

CBRN-CADS v3 does not replace existing Korean institutional reporting systems. It generates the reports they require, with metadata pre-populated:

• KAHIS (Korea Animal Health Integrated System) — for any incident involving zoonotic pathogens (H5N1, anthrax, brucellosis). CARE module auto-generates the KAHIS Form A and routes to the Animal and Plant Quarantine Agency.
• Olbaro (the Korean waste management tracking system) — for any incident generating CBRN-contaminated waste under the Waste Management Act Enforcement Decree Annex 2. The system pre-populates the Olbaro transfer manifest with chain-of-custody documentation.
• ICAO Annex 14 Wildlife Strike Report — for any bird-strike or wildlife event at an airport or vertiport. The system auto-generates the IBIS-compliant report with weather metadata, location coordinates, and casualty-of-aircraft assessment.

Why This Architecture Matters

In a textbook simulation, a CBRN incident response is a sequential 11-step process. In a real urban event, all 11 steps run in parallel under time pressure, with information arriving asynchronously from multiple sensors and human reporters. The traditional command-and-control architecture struggles to coordinate this without a digital backbone.

CBRN-CADS v3's contribution is the parallel decision tree: each of the 11 nodes runs independently with AI-recommended outputs, the four-agency coordination layer (NIS, Ministry of Defense, National Police, Ministry of Environment) sees the same operating picture, and the after-action report is built incrementally as decisions are made — not reconstructed afterward from incomplete logs.

What Validates the Architecture

The technical lineage is explicit. CBRN-CADS v3 inherits from the NDMI 2013 government pilot work that validated six dispersion models and Pasquill-Gifford coefficients for Korean atmospheric conditions. The copyright registration (Certificate of Copyright Registration No. C-2026-024315, May 20, 2026, Korea Copyright Commission) covers the software architecture, the decision logic, and the integration interfaces.

For Korean defense, public safety, and civil-aviation organizations evaluating CBRN decision support systems, CBRN-CADS v3 offers a documented architecture with explicit government and academic lineage — not a black-box product, but a transparent decision tree that any responder can audit during or after an event.

---

UAM Korea Tech Co., Ltd. — CBRN-CADS v3 specification available under NDA to qualified parties.