SAFERS

Project Description

SAFERS takes action to demonstrate how a data-driven, AI-powered, and community-connected approach to emergency management can strengthen society’s resilience against forest fires. As climate change drives more frequent and more extreme wildfire events, the project proposes a call for action — coupling heterogeneous Big Data, advanced predictive models, and Artificial Intelligence to support first responders, citizens, and decision makers across the whole emergency management cycle.

SAFERS develops, tests, and demonstrates a comprehensive Emergency Management System (EMS), complemented by data, models, methods, and tools able to forecast risk, detect and track active fires, and estimate impacts, protecting European territories from the growing threat of forest fires. Earth Observation data from Copernicus and GEOSS form the primary data source, combined with crowdsourced information from social media, in-situ smoke and fire sensors, and dedicated mobile applications used by both citizens and professional first responders.

The system acts along the three key phases of the emergency cycle: generating risk maps and early warnings in the preparedness phase, estimating fire extension and propagation based on forecasted weather and soil conditions during the response phase, and computing post-event impacts and soil recovery in the recovery phase.

LINKS Foundation Activities

LINKS Foundation participates as project coordinator, leading the design of the overall system architecture and the development of the platform’s AI-based intelligent services, geospatial data infrastructure, and crowdsourcing tools.

AI-Based Early Warning & Decision Support Services

Coordination of the operational Intelligent Services (IS), providing autonomous, periodical updates for specific data streams, including:

• Sub-seasonal Weather Forecasts
• Operational Early Warnings
• In-situ Smoke and Fire Detection Systems
• the Decision Support System

On-Demand Wildfire Mapping & Forecast Services

Development of on-demand services triggered by user-defined map requests, retrieving and processing data to generate:

• EO-based fire delineation
• Post-fire monitoring outputs
• On-demand wildfire forecasts

Once results become available, the remaining services are notified through the message bus, enabling the required follow-up actions.

Social Media & Chatbot Crowdsourcing

Design of two crowdsourcing solutions for direct and indirect citizen engagement:

• a chatbot enabling structured, geolocated, and multimedia data collection from citizens and field agents
• a social media module that gathers and classifies real-time Twitter posts, extracting emergency events and estimating their impacts

Geospatial Data Management & Interoperable Web Platform

Design and delivery of an INSPIRE-compliant data layer acting as the platform’s central storage system, composed of:

• a GeoData Repository (GDR) for heterogeneous raw geospatial data
• Importer and Mapper modules generating OGC-compliant WMTS layers on the tile server
• a central backend, serving as the communication hub between users and services
• a web-based dashboard visualizing WMS/WMTS layers, social media data, and real-time in-situ camera feeds

Funding & Dates

SDGs — Sustainable Development Goals

🌍 Goal 13 — Climate Action

Key Technologies

🤖 AI & Machine Learning for Wildfire Risk

Algorithm-based models estimating forest fire risk, generating early warnings, forecasting fire extension and propagation as a function of weather and soil conditions, and computing post-event economic and environmental impacts.

🛰️ Copernicus & GEOSS Earth Observation

Exploitation of satellite Earth Observation data from Copernicus and GEOSS as the platform’s primary data source, feeding both operational and on-demand Intelligent Services.

📡 In-Situ Smoke & Fire Detection Sensors

Integration of in-field smoke and fire detection sensor networks, providing real-time, ground-level data that complements satellite observations.

💬 Crowdsourcing Chatbot & Social Media Analytics

A conversational chatbot for structured, geolocated, multimedia data collection, paired with a social media module that classifies real-time Twitter posts to detect and estimate the impact of emergency events.

🗄️ Geospatial Data Repository & OGC/INSPIRE Interoperability

An INSPIRE-compliant data layer built around a GeoData Repository, with Importer and Mapper modules generating OGC-compliant WMTS layers for standards-based interoperability.

🖥️ Decision Support System & Web Dashboard

A central backend hub connecting users and services, paired with a web-based dashboard for intuitive visualization of geospatial layers, social media data, and real-time in-situ camera feeds.

Open Source Resources

The following core services are openly available on GitHub:

• SAFERS Dashboard Frontend
• SAFERS Dashboard Backend
• SAFERS GeoData Repository
• SAFERS Importer & Mapper
• SAFERS Message Broker

Technical Documentation

• SAFERS - Technical Report
• SAFERS - Chatbot User Guide
• SAFERS - Dashboard User Guide

Project Partners

Project Information

• 🇪🇺 SAFERS on CORDIS
• 🌐 SAFERS Website
• 🐦 SAFERS on X
• 💻 SAFERS on LinkedIn

For further information or commercial support, please contact us at contact@safers-project.eu.