Preventing fire events by rediscovering and extending deep learning methods - PREFERRED

preferred

 

Project Number: ATTP4-0340489
Project Acronym: PREFERRED

General information
Project title: Preventing fire events by rediscovering and extending deep learning methods
Starting date: 04/08/2022
Duration in months: 17

Funding scheme: Regional Operational Programme of Attiki
Call (part) identifier: Research and Innovation Synergies in the Region of Attica
Topic: DIGITAL-2022-CLOUD-AI-02-OPEN-AI: LC-GD-1-1-2020 Public Sector Open Data for AI and Open Data Platform
Keywords: Climate change, fires, dynamic platform, geospatial analysis, forecast, urban ecosystems, Machine Learning, Deep learning, Earth Observation, Big Data

Website: https://www.preferred.gr/

preferred founded 

The escalation of climate change, all over the world, increases the vulnerability of forests to fires, and threatens their very existence as well as, especially in the wildland urban interface (WUI) areas, the life and safety of citizens, social cohesion and prosperity. The plethora of recorded events with many victims and severe destruction of infrastructures & properties, highlights the danger of the environment we live in and the critical need for the most effective management of events through informed decisions (timely & valid information).

The proposed development aims to offer innovative services to organizations directly or indirectly involved in forest fires management; Institutional Bodies responsible for crisis management (fire brigades, Civil protection Bodies, etc.) and / or bodies managing specific sites, environmental parks, archeological sites, etc.

Focusing on Priority 4: Enhancing disaster preparedness, of the Sendai Framework for Disaster Risk Reduction and in Urban Areas (WUI), the project objectives are developed in four axes:

A- Development of a direct access platform for stakeholders

B- Event Management Needs Coverage: Preparedness - Impact Mitigation & Crisis Management

C- Provision of free data and open-source software

D- Development of innovative deep learning methods to solve the problem of fire risk assessment at the next day level

The operation of the web application is expected to drastically improve the efficiency of fire incident management in urban areas, increase citizens' sense of security, and reduce management costs (mobilization, compensation, etc.).