UN EW4All 2027 — Time is Running Out

Disasters happen.
Is your community ready?

Time Remaining to UN EW4All 2027 Deadline

— Days
— Hours
— Minutes
— Seconds

DEADLINE: 31 DECEMBER 2027 · UN GENERAL ASSEMBLY RESOLUTION A/RES/78/210

SafeGround is building a resilient early warning infrastructure — designed to function where no network exists. Open source. Infrastructure-independent. Aligned with the UN Early Warnings for All 2027 mandate.

Download White Paper Partner / Collaborate →
3.5B
People without access to multi-hazard early warning systems (WMO, 2024)
2×
Increase in extreme weather events over the last 40 years (WMO)
10:$1
Return on early warning investment — every $1 saves at least $10 in disaster response
~50%
Of UN member states have adequate multi-hazard early warning systems — as of 2024

Where This Began

Hurricane Sandy showed us the gap. We've been building the bridge ever since.

In October 2012, as Sandy bore down on the New Jersey coastline, a network of disaster experts — convened through the EdgeKnowledge think tank and the Friendship Train Foundation — realised something that would take years to fully act on.

The warnings existed. The knowledge existed. The problem was infrastructure: when Sandy hit, communications failed precisely when they were needed most. In the immediate aftermath, long-range radios became the team's lifeline — and the seed of what would eventually become SafeGround was planted.

PROJECT FOUNDER & LEAD — DR. FRANCES KIERNAN
Emergency Medicine Physician · NHS Innovation Fellow · Former Technology Lead, GBREX & Monmouth County Long-Term Recovery Group

Read the Full Story →
Hurricane Sandy satellite imagery — the storm that catalysed SafeGround
Hurricane Sandy, October 2012 · The storm that changed everything

The Gap We're Closing

The technology exists.
The deadline is 2027.
Billions are still unprotected.

The UN Early Warnings for All initiative represents the most ambitious commitment to disaster risk reduction in history. Yet the gap between commitment and capability has never been wider — because most early warning systems depend on the same infrastructure that disasters destroy first.

📡

Infrastructure Dependency

Cellular networks, power grids, and internet infrastructure are the first casualties of major disasters. Existing early warning systems rely on them — and fail precisely when they are needed most.

🌍

Last-Mile Coverage

The communities most at risk — remote, rural, low-income — are the least served by commercial infrastructure. Yet they face the highest exposure to earthquakes, floods, landslides, and wildfires.

⏱️

Warning Time

A warning that arrives after power has failed, or that reaches a community hub but not individual households, is not a warning system. SafeGround is designed from first principles for the last mile.

Beyond Natural Disaster

Early warning for every threat that demands a response

The SafeGround architecture — sense, assess, classify, alert, escalate — was designed for natural hazards. The same framework applies wherever communities face threats that require real-time detection, triage, and coordinated response.

🌊

Natural Hazards

Seismic, flood, wildfire, storm surge, landslide, volcanic. The original mandate. Sensor networks feeding AI-assessed threat briefings to communities before infrastructure fails.

🧪

Chemical & Industrial

HAZMAT incidents, industrial accidents, toxic releases. Distributed sensor arrays and anomaly detection for communities near high-risk facilities or transport corridors.

🦠

Biological

Outbreak detection and sentinel case escalation. The same sense-assess-classify architecture used for natural disaster applies directly to public health surveillance and early epidemic warning.

☢️

Nuclear & Radiological

Radiation monitoring and escalation for communities near nuclear facilities. Infrastructure-independent alert networks that function even when grid power is lost.

The physician's perspective — trained in emergency medicine, disaster response, and biosecurity — shapes every layer of the SafeGround threat architecture. Read the research →

Earth from space — the SafeGround system is designed for global coverage
Infrastructure-independent. Works where nothing else does.

The SafeGround Solution

LoRa mesh networking. No internet required. No power grid required.

SafeGround uses LoRa (Long Range) radio technology to create self-healing mesh networks that operate entirely independently of cellular, internet, or grid infrastructure. Nodes relay signals across distances up to 15 km, forming a community-owned early warning backbone that survives the very disasters it protects against.

✓ Three-tier architecture: User Node, Contributor Node, Admin Hub
✓ CAP v1.2 alert format, ATAK & What3Words integration
✓ XChaCha20-Poly1305 end-to-end encryption
✓ AI-driven threat intelligence (seismic, fire, weather, flood)
✓ Open source — built for collaboration
Explore the Technology →

Proof of Concept

SENTINEL: Where the System Lives

Our PoC device collapses the full three-tier SafeGround architecture into a single deployable unit. It monitors, assesses, and broadcasts — autonomously.

HOME Mode

Static seismic and environmental monitoring. AI threat assessment every 15 minutes. Continuous display.

ALERT Mode

Triggered on threshold breach. Seismic >3.5 local, barometric drop >8hPa/3h. Broadcasts mesh + email + push.

GO Mode

Portable, battery-powered. GPS beacon, mesh communicator, last-known threat briefing on e-ink display.

Electronic hardware components — the SENTINEL device uses LoRa, GPS, and Raspberry Pi hardware
SENTINEL hardware: LILYGO T-Beam Supreme · Raspberry Pi 4 · RAK2287 LoRaWAN · ADXL345 Seismic

Pissouri, Cyprus — sitting on the African-Eurasian plate boundary, with documented landslide risk and growing wildfire exposure, and zero existing LoRa infrastructure — is our candidate PoC site. The ideal proving ground.

Technical White Paper

SafeGround EWS: A Blueprint for Infrastructure-Independent Disaster Communications

Full technical specification, origin narrative, architecture deep-dive, EW4All alignment analysis, and implementation pathway for pilot deployment.

Read & Download

The Mandate

Every person on Earth.
By 2027.
Not a goal. A commitment.

In 2022, the United Nations launched Early Warnings for All — the most ambitious disaster risk initiative in history. Its target: universal coverage of multi-hazard early warning systems by the end of 2027. As that deadline approaches, the gap between commitment and reality is stark.

SafeGround is not waiting for infrastructure investment to reach the most vulnerable. We are building the infrastructure itself — open, distributed, and designed to function in the absence of everything else.

Explore EW4All →
Emergency responders — early warning saves lives across every community
Every minute without warning is a minute that communities cannot prepare

Stay Informed

Join the SafeGround Network

Get updates on project development, pilot deployments, open source releases, and opportunities to collaborate.

No spam. Infrequent updates only. Unsubscribe any time.

Open Source

Built to be shared. Designed to be replicated.

SafeGround is an open source project. Every protocol, firmware, and architecture decision is available for adaptation, improvement, and deployment by NGOs, researchers, governments, and communities worldwide. The goal is not one system — it is a standard.

MIT Licence Open Hardware CAP v1.2 UN SDG 11 & 13
View on GitHub

Research & Output

From the Field

Project publications, technical updates, and research from the SafeGround team.

Publication May 2026

Before the Storm: Early Warning as a Clinical Imperative

SafeGround, LoRa mesh networks, and the physician's role in disaster intelligence. Early warning systems are almost universally framed as an engineering problem. They are not. They are a clinical one.

Dr. Frances Kiernan · Substack Read article →
Publication April 2026

Genomic Dual-Use Risk and the Weaponisation of -Omic Technologies

How convergent AI and genomic advances have transitioned biosecurity threats from "possibly plausible" to probable — and what the SafeGround threat-intelligence architecture demonstrates for AI biosecurity work.

Dr. Frances Kiernan · Substack Read article →
View All Research & Output →