CBRN Hazmat as Lattice Entities: Schema Design for CADS Integration

Abstract

The Anduril Lattice platform has emerged as one of the most consequential autonomy integration frameworks in Western defense since DARPA's early autonomous systems programs. Its Entity schema — a structured, extensible data model for representing any trackable battlefield object — was designed primarily around kinetic threats: vehicles, aircraft, personnel, and weapon systems. Yet the most lethal threats on the modern battlefield are often invisible to conventional ISR: nerve agent plumes, biological aerosols, and radiological point sources. This article argues that UAM KoreaTech's CBRN-CADS multi-sensor detection platform is uniquely positioned to publish real-time chemical and biological detections as first-class Lattice Entities, using the platform_type: Animal+ classification and a purpose-designed Hazmat extension payload. The resulting integration closes the detect-to-decontaminate loop within a single operational environment — with BLIS-D's waterless 90-second decon cycle executable as a Lattice-tasked action. This article presents the schema design rationale, the AbriIndex urgency metric, and the strategic case for making CBRN hazmat sources visible in the Lattice common operating picture before the next mass-casualty event forces the issue.

---

1. Historical Anchor — The Tokyo Subway Sarin Attack, 1995

Inner Landscape

On the morning of 13 March 1995, the Aum Shinrikyo cult released Sarin on five Tokyo subway lines during peak commuting hours. The operational commanders responsible for the response — Tokyo Metropolitan Police, Japan Self-Defense Forces NBC units, and emergency medical services — operated from profoundly fragmented situational awareness. Each responder organization maintained its own communication channel. There was no shared common operating picture. Casualty reports from Kasumigaseki station could not be automatically correlated with atmospheric drift data from Kodemmacho. Decisions about evacuation corridors were made on voice radio with minutes-old information. The inner landscape of those commanders was one of disciplined competence paralyzed by information isolation — each unit knew its immediate environment but was blind to the system-wide chemical threat geometry.

Environmental Read

The environmental factors that amplified the attack's lethality were precisely those that a modern Lattice-integrated CBRN architecture would mitigate. Tokyo's subway ventilation system created unpredictable airflow that dispersed Sarin beyond the initial release points; responders had no real-time atmospheric dispersion model. The 13 stations ultimately affected spanned multiple command jurisdictions. First responders, lacking chemical detection gear, became secondary casualties: 135 emergency service personnel were affected, reducing the response capacity at the moment it was needed most. The lesson the environmental read teaches is not that the responders lacked courage or training — it is that they lacked a shared data layer that could have georeferenced the hazmat source, projected its drift, and automatically allocated decontamination assets.

Differential Factor

What made Tokyo 1995 different from prior chemical incidents was not the agent or the scale — it was the urban, networked infrastructure context. The attack was deliberately designed to exploit the connectivity of a modern city: a subway system that moves hundreds of thousands of people through a small geographic space every hour is also an aerosol distribution network. The differential factor was the absence of any sensor-to-command data pipeline. A single IMS detector at Kasumigaseki could have, in principle, triggered an automated station lockdown and ventilation reversal within 90 seconds. The infrastructure existed. The sensor-to-command integration did not. This is precisely the gap that the CBRN-CADS → Lattice Entity pipeline is designed to close — not just in subway stations but on military platforms, forward operating bases, and any Lattice-connected node.

Modern Bridge

The operational gap exposed in Tokyo is structurally identical to the gap that exists today on Lattice-connected battlefields: CBRN threats are real, but they are invisible to the autonomy stack because no sensor has formally published them as trackable entities. UAM KoreaTech's integration work directly addresses this. By modeling each CBRN-CADS detection event as a Lattice Entity — with a georeferenced origin point, a drift vector derived from onboard meteorological sensors, and an AbriIndex urgency score — commanders gain the same situational awareness over chemical threats that they currently have over enemy vehicle formations. The bridge from 1995 to 2026 is the Entity schema itself.

---

2. Problem Definition — The CBRN Sensor-to-Command Gap

The global CBRN defense market was valued at approximately USD 16.3 billion in 2023 and is projected to reach USD 22.1 billion by 2028, growing at a CAGR of 6.3%, according to MarketsandMarkets. Yet the majority of that investment flows into detection hardware and personal protective equipment — not into the software integration layer that connects sensor outputs to command decisions. A 2022 NATO STO technical review identified C2 integration of CBRN sensor data as one of the top three capability gaps across Alliance ground forces. Specifically, it found that fewer than 12% of deployed CBRN detection systems in NATO exercises were capable of pushing structured detection data to a brigade-level common operating picture in real time.

The consequences are measurable. In NATO's CWIX 2023 interoperability exercises, CBRN sensor data was transmitted to command nodes with an average latency of 6.8 minutes — during which a Sarin plume traveling at 3 m/s in a 10 km/h wind covers approximately 250 meters of contested space per minute. The current NATO STANAG 2112 reporting standard provides a structured message format for NBC warnings, but it was designed for voice radio and structured text — not for ingestion by an autonomous tasking system like Lattice. There is no native mechanism for a STANAG 2112 NBC-1 report to instantiate a tracked Entity in Lattice's entity graph. That is the specific technical gap UAM KoreaTech's schema work addresses.

---

3. UAM KoreaTech Solution — CBRN-CADS as a Lattice Entity Publisher

CBRN-CADS is a multi-sensor fusion platform combining ion mobility spectrometry (IMS), Raman spectroscopy, gamma/neutron detection, and quantitative PCR for biological agents. Its AI inference layer produces a classified detection event within under 45 seconds of agent exposure. The schema integration layer developed by UAM KoreaTech translates each confirmed detection event into a Lattice Entity object with the following core fields:

• entity_id: UUID generated from sensor node ID + timestamp
• platform_type: Animal+ (bridging classification pending Lattice Hazmat type)
• aliases: human-readable label (e.g., "Sarin_Source_Grid_37TDE_4421")
• provenance: CBRN-CADS v2.4 / IMS+Raman fusion
• TEMPLATE_TRACK: georeferenced track history with 10-second update intervals
• hazmat_extension: agent class, CAS number, confidence score (0–1), concentration in mg/m³, wind drift vector (bearing/speed), exclusion radius in meters, and AbriIndex score (0–100)

The AbriIndex is UAM KoreaTech's proprietary urgency metric, computed from agent lethality class (OPCW Schedule 1/2/3), measured concentration relative to IDLH thresholds, and time-to-lethal-exposure at current drift rate. An AbriIndex above 75 triggers an automatic shelter-in-place recommendation to all Lattice nodes within the projected hazard polygon.

Once the Entity is live in the Lattice COP, BLIS-D decontamination systems linked to the same Lattice environment receive an automated tasking packet: agent type, recommended neutralization protocol, and affected platform list. BLIS-D's bleed-air dry decon cycle — requiring no water, no consumable chemical stockpile, and completing in 90 seconds — then executes against the designated platforms, with completion status reported back as an Entity attribute update.

---

4. Strategic Context — Why Korea, Why Lattice, Why Now

The Korean Peninsula presents a uniquely concentrated CBRN threat environment. The Republic of Korea's Ministry of National Defense has publicly assessed that the DPRK maintains chemical weapons stockpiles estimated between 2,500 and 5,000 metric tons, including Sarin, VX, and mustard agent. This is the world's third-largest chemical weapons program by volume, according to the IISS Military Balance. Any conventional conflict on the peninsula would almost certainly involve chemical employment within the first 72 hours of hostilities — making sub-minute sensor-to-command CBRN data pipelines an existential operational requirement, not a procurement nicety.

Simultaneously, the U.S. Indo-Pacific Command's accelerating adoption of Anduril Lattice as its preferred autonomy integration framework — evidenced by the 2023 REPLICATOR initiative and subsequent INDOPACOM AI contracting activity — means that any Korean defense capability seeking interoperability with U.S. forces must speak the Lattice Entity language natively. UAM KoreaTech's dual-use positioning — NATO STANAG 2112 compliant on the structured reporting side, Lattice Entity publisher on the autonomy integration side — makes CBRN-CADS the only Korean-origin CBRN sensor with a credible claim to full-stack U.S. force interoperability.

Regulatory tailwinds reinforce this positioning. The Korea Defense Acquisition Program Administration (DAPA) K-CBRN modernization program, with a planned budget of approximately KRW 340 billion through 2030, explicitly requires AI-enabled sensor fusion and C2 integration as mandatory capability thresholds. European NATO allies, post-Salisbury and post-Ukraine, are similarly accelerating CBRN C2 integration investment under NATO's CBRN Defense Roadmap 2030.

---

5. Forward Outlook

UAM KoreaTech's CBRN-CADS → Lattice Entity integration roadmap targets the following milestones across the next 18 months:

Q3 2026: Schema specification v1.0 published as open technical reference, submitted to NATO CBRN C2 working group for STANAG alignment review. Hazmat extension payload frozen for first integration test.

Q4 2026: Live fire integration exercise with a Lattice-connected UAS platform, demonstrating sub-60-second Entity publication latency from CBRN-CADS detection event.

Q1 2027: BLIS-D automated tasking loop validated in a NATO CWIX interoperability exercise environment, demonstrating end-to-end detect-to-decontaminate closure under 3 minutes.

Q2 2027: AbriIndex metric submitted for independent validation against OPCW reference scenarios. Integration package available to Allied defense integrators as a licensable SDK.

The commercial pathway runs in parallel: UAM KoreaTech is in active discussions with two Tier-1 prime defense contractors regarding CBRN-CADS as a sensor payload for Lattice-integrated UAS platforms currently in INDOPACOM evaluation. The Hazmat Entity schema work positions the company as the domain-specific CBRN data layer within what is rapidly becoming a Lattice-centric autonomy ecosystem.

---

Conclusion

The Tokyo subway attack of 1995 killed 13 people and injured more than 5,000 because a lethal invisible agent moved faster than fragmented human command structures could respond. Thirty years later, the technology to publish that Sarin source as a real-time, drift-tracked, autonomy-readable Lattice Entity exists — and UAM KoreaTech's CBRN-CADS + BLIS-D integration is the architecture that closes that gap. Making CBRN hazmat sources first-class citizens of the Lattice common operating picture is not a product feature; it is the operational imperative that Tokyo 1995 wrote in blood and that no Alliance commander should face without in 2026.