Skip to main content
AiABy LyDian
Operations liveMonitoring 127 active incidents across 81 regions

National-scale disaster coordination, built for zero-trust operations.

The command surface that holds the ground line when the grid goes dark.

AiA fuses real-time incident triage, team dispatch, long-range radio/short-range mesh communication, seismic telemetry, and family-reunion logistics into a single operator-first platform — designed to keep running when cell towers, fibre, and power do not.

AiABy LyDian
Operations live
USGS · 11 adapters live
Mesh · readyUplink · LoRa + Sat
  • Incident triage in under 90 seconds
  • Mesh-first communications with satellite fallback
  • End-to-end encrypted operator channels
  • KVKK · GDPR · ISO 27001 controls built-in
  • National scale, field-ready, sovereign-hosted
Operating envelope

Built to hold the line when seconds cost lives.

Median triage time
<90s
From SOS trigger to dispatched responder
Always-on operations
24/7
Grid-independent, solar-backed nodes
Multi-transport fabric
Mesh + long-range radio
short-range · peer-to-peer Wi-Fi · long-range radio · Satellite
Coordinated domains
16
Incidents · Teams · Comms · Seismic · ICE · Reunion · Logistics · Ops
Platform capabilities

Every capability your incident commanders need, and none they don't.

Every module is production-grade, built around the the service core core, the datastore with, the queue fabric queues, a the encryption core over internal RPC, and the realtime fabric realtime. No mocks. No placeholders. Only capability.

01

Incident Triage Engine

Geotagged SOS, accident, medical, and fire reports routed through severity-aware state machines (PENDING → ASSIGNED → RESPONDING → RESOLVED) with full audit trail.

Every incident carries a canonical state model, an immutable activity log, response SLA counters, operator attribution, and WebSocket fan-out. Severity (RED / YELLOW / GREEN / BLACK) drives automatic team pre-assignment suggestions.
02

Team Dispatch & Availability

Five-type responder fleet (MEDICAL · FIRE · POLICE · RESCUE · SUPPORT) with live capacity, geofenced availability, and bidirectional dispatch.

Capacity, equipment load-out, last-seen geolocation, and radio identity travel with every team. Dispatch creates an atomic pairing with the incident, hardware-signed and replayable for after-action review.
03

Mesh + long-range radio Communications

short-range mesh, peer-to-peer Wi-Fi, long-range radio long-range, and field nodes relay firmware keep operator and citizen traffic alive when infrastructure fails.

Store-and-forward delivery with eventually-consistent merge convergence, long-range long-range wristbands for passive triage, and a solar-backed relay kit BOM that is field-deployable in under 14 minutes. Encrypted end-to-end with authenticated encryption over the the encryption core.
04

Seismic Telemetry & Early Warning

AFAD and Kandilli feeds ingested, deduplicated, and enriched with fault-line context, heatmapped on a real-time 3D mapping 3D globe.

Every event is cross-validated across agencies, assigned a canonical UAA-ID, and pushed over the WebSocket fabric. Fault-line overlays, building-stock inspection triggers, and fire-risk co-factors are computed server-side in under one second.
05

Family Reunion & ICE Profiles

Privacy-first reunion queue and In-Case-of-Emergency profile vault — citizen-owned, operator-read-only, KVKK-compliant.

Profiles carry blood type, chronic conditions, dependents, pet information, and trusted contacts. Family reunion matches operate on hashed identifiers; a match surfaces only to authorised reunion operators with audit-logged consent.
06

Resource & Shelter Inventory

Live positions and capacity of shelters, field hospitals, water points, fuel, medical stock, and generator fleet.

Inventory cursors update in real time through operator ingest or partner agency webhooks. A single query returns the closest viable resource with traffic-aware ETA and residual capacity.
07

Realtime Fabric

the realtime fabric gateway on a dedicated port fans incident, team, comms, seismic, and system events to every authorised surface.

Rooms per operator, per region, per incident. Backpressure-aware, with automatic reconnection and offline queue replay. Cross-device consistency via eventually-consistent merge-merged state.
08

Sovereign Encryption Core

A isolated service (authenticated encryption) handles every field-sensitive envelope — operator messages, ICE profiles, reunion matches.

Keys rotate through a hardware-root module. The core binary is reproducible-built, signature-pinned, and runs in a network-isolated sidecar. Compromising the web layer does not compromise the ciphertext.
09

Offline-First & Downloads

Operators can pull mapsheets, ICE bundles, incident snapshots, and team manifests for offline use during connectivity brownouts.

Every download is signed; every upload is idempotent with an offline replay queue. The operator console keeps working for up to 72 hours without backhaul.
10

Full-Stack Observability

metrics, logs, dashboards, and -tracked database change feeds give operations leads ground-truth on the platform itself.

SLOs: 99.95% gateway availability, <500 ms incident-create p95, <2 s WebSocket propagation p99. Every controller, queue, and internal RPC edge emits structured traces.
11

AI Early Warning

The moment an earthquake is detected, a second-by-second countdown begins before S-waves reach your cities; risk scores are calculated based on fault proximity, soil type, and population density.

Scenario: A 6.8 magnitude earthquake in Marmara. The system detects the P-wave in 4 seconds, issues a warning 23 seconds before it reaches Istanbul. Simultaneously calculates landslide, tsunami, and liquefaction probabilities, displaying the cascade hazard chain on screen.
12

Cascade Hazard Analysis

An earthquake is never just shaking — automatically detects subsequent tsunami, landslide, flood, liquefaction, and fire risks with probability percentages.

Scenario: A 7.1 earthquake in Hatay. The system calculates landslide risk in the Amanos Mountains at 67%, tsunami risk in the Gulf of Iskenderun at 34%, and flood risk in the Ceyhan river basin at 22%, displaying the entire chain on a single screen.
13

Multi-Channel Alert System

Creates disaster warnings in international CAP standard; delivers simultaneously to all citizens in the target area via instant notification, SMS, email, and mobile push.

Scenario: Flood warning in the Black Sea region. The system creates alerts in Turkish, English, Arabic, and Kurdish, delivering to all devices within the polygon drawn on the map through 4 channels simultaneously. Delivery status is tracked in real-time.
14

Damage Assessment

AI-powered damage classification from citizen reports, drone imagery, and satellite data — building-level assessment as undamaged, minor, major, or destroyed.

Scenario: Within the first 2 hours after the earthquake, 847 citizens submit photo damage reports. The system classifies each report with AI, cross-validates with satellite imagery, and creates a building-level damage layer on the map. Field teams are prioritized to areas with the most destruction.
15

Digital Twin Simulation

Simulates earthquakes that haven't happened yet — building damage, casualty estimates, evacuation routes, and economic loss calculations for disaster preparedness planning.

Scenario: The 'Istanbul Marmara 7.5' preset scenario is run. The system calculates damage distribution for 42,000 buildings, estimates 2,800 casualties and 180 billion TRY economic loss. Suggests five evacuation routes with estimated duration for each.
16

Live Disaster Data Stream

AFAD, Kandilli, USGS earthquake data, GDACS international alerts, and NASA fire detections — all updated in real-time every 30 seconds.

Scenario: At 03:14 AM, a 5.2 earthquake in Van. AFAD delivers data in 30 seconds, Kandilli in 45 seconds, USGS in 90 seconds. Data is automatically cross-validated, duplicates eliminated, and reflected on the command screen as a single verified event.
Scenarios, drawn from reality

What the platform actually does, minute by minute.

These are not marketing vignettes. They are the operational flows the platform executes today against the real backend — real incidents, real teams, real telemetry.

Major earthquake · Turkey timeline

7.2 Mw strike on the Eastern Anatolia corridor

  1. T+00:00
    AFAD telemetry ingested

    P-wave detection triggers automatic triage. AFAD confirms M7.2 at 10km depth near Erzincan. System calculates S-wave ETA for 20 cities.

  2. T+00:03
    Cascade analysis fires

    AI detects 45% landslide probability in surrounding mountains, 12% liquefaction risk in alluvial plains. Three adjacent provinces flagged.

  3. T+00:12
    Citizens begin self-reporting

    340 geo-tagged SOS reports in first hour. Severity triage classifies 47 as RED. USAR teams auto-dispatched to highest-density collapse zones.

  4. T+02:30
    Family reunion queue saturates

    15,000 missing-person lookups. 2,400 matched via hashed identity tokens. Consent-gated disclosure protects privacy while enabling reunion.

  5. T+12:00
    After-action materialises

    Full incident timeline, resource utilization, SLA compliance, and shelter capacity reports generated automatically for AFAD review.

Infrastructure brownout timeline

72-hour grid and backhaul outage, remote province

  1. Day 1
    The mesh takes over

    Operator handsets form a short-range, peer-to-peer Wi-Fi mesh. Field nodes relay firmware already flashed onto deployed units and reattach to a long-range radio backbone. Incident reporting continues without a pause.

  2. Day 1 · evening
    Solar relay kits drop in

    Pre-positioned solar relay kits (bill of materials published) are handed to volunteer teams. Each kit holds a full day of traffic in its offline queue and replays it on reconnection.

  3. Day 2
    The encryption core stays sovereign

    Even in degraded conditions the encryption core keeps sealing every field-sensitive envelope. No message leaves a node in the clear.

  4. Day 3
    Backhaul returns

    The moment fibre or satellite uplink comes back, every queued event replays in causal order. Statistics are reconciled; no incident is lost.

Mass-casualty incident timeline

Motorway pile-up

  1. T+00:00
    First SOS received

    A citizen triggers an SOS from the mobile app. The incident enters PENDING state at RED severity, geotagged to three-metre accuracy.

  2. T+00:01
    ICE profile attaches automatically

    The citizen's ICE profile — blood type, allergies, chronic conditions — is attached with an audit entry; visibility is restricted to the assigned MEDICAL team alone.

  3. T+00:04
    Dispatch completes

    Dispatch matches the incident to the nearest MEDICAL team in AVAILABLE state. Team status moves to DISPATCHED; the estimated time of arrival is published to the operator and to next of kin.

  4. T+00:22
    Arrival and handover

    Team status moves to ON SCENE. The operator can log clinical notes, patient count and transport destination; every entry is encrypted at rest by the encryption core.

Citizen mobile

One app. Two devices. Four million citizens.

The operator console, in your pocket — offline-first, signed end-to-end, available in eleven languages.

  • Offline-first BLE + LoRa mesh — keeps working when towers fall.
  • Signed, end-to-end encrypted SOS — only authorised responders can read.
  • ICE profile stored on the device — never leaves your phone unencrypted.
  • Eleven languages, voice command, large-print mode for emergencies.
iOSComing soon — TestFlight betaAndroidDownload APK · 24 MB

v1.0.0 · 24 MB · Android 8+ · iOS 15+

Who it is for

One platform, many hands on the wheel.

Operators & dispatchers

Triage incidents, move teams, run the mesh, and answer family reunion calls with every data point surfaced at the right moment.

Field responders

A glove-friendly, high-contrast mobile console with mesh fallback, offline queue, and push-to-record incident flows.

Operations leads

A single pane for regional situational awareness, SLA and workload analytics, and real-time risk heatmaps.

National civil-protection authorities

API-first integrations with AFAD, Kandilli, municipal services, and partner agencies via webhooks and signed data exports.

Citizens

A mobile app that lets people report, reunite, and receive early-warning alerts in their own language, over any network available.

Frequently asked

Before you deploy.

It is production software. Sixteen backend controllers, a -managed the datastore schema, a encryption sidecar, a the realtime fabric realtime fabric, a the operator console operator console, and a the mobile client field client all ship today.

Put it in the field

A platform this serious deserves a serious onboarding.

Book a live walkthrough with the engineering team. We will stand up a scoped operator console, route a simulated incident, and show the mesh reconvening after a synthetic brownout — end to end, on your hardware.