The National Disaster-Warning Case Study on heain-core Validating the HEAIN Framework for Warnings that Work when Cut Off and Can Be Audited, with Real-Event Replays and Scale Tests up to 872 Nodes

Abstract Thailand can already send warnings to mobile phones: a nationwide Cell Broadcast test on 20 January 2026 reached handsets within about 10 seconds [25]. Recent disasters show that the failures lie before the send button: who holds the data, who may press it, and how long the decision waits. The earthquake warning of 28 March 2025 reached the telecom regulator 82 minutes after the quake [26]; the Hat Yai evacuation order came about 63 hours after an upstream station reached Extreme [28]; and the first Cell Broadcast of the 2026 Bangkok floods went out a day after the flooding began [31]. This is Volume 1 (Overview) of the third HEAIN case study: a hypothetical four-tier national warning system, from the Department of Disaster Prevention and Mitigation (DDPM) down to village Micro Nodes. The author turned the design [15] into a 29-node simulation running the real heain-core on one HP Z840, tested 26 scenarios (H1–H26) in 8 rounds, and then replayed 4 real events (E1–E4): the Hat Yai flood of 2025, the Bangkok floods of 2026, the earthquake of 28 March 2025, and the 2026 PM2.5 season. Each event ran twice on a layout of 34–39 nodes, "as it happened" and "with HEAIN", against 40 requirements (R1–R40) drawn from what failed. In the final run every scenario and every event passed all 502 checks, with no failure and no gap left open. A 24-hour soak under 69 random faults delivered 109,734 jobs, each exactly once, and the scale tests ran Bangkok in full (252 nodes) and 872 of 1,006 national tier heads. It led to 8 rounds of core changes (C1–C8), each closing a gap the scenarios or events exposed. With HEAIN the Hat Yai evacuation order went out 1.1 hours after the station reached Extreme instead of 62.9 hours, and the first earthquake CAP went out in minute 2 instead of minute 85. Keywords: HEAIN Framework, heain-core, Case Study, Disaster Early Warning, Cell Broadcast, CAP 1.2, Corroboration, Network Partition, Transparency Log, Scale Testing, Edge AI