Advance-ALERT is an AI-powered module for the European's EU-ALERT disaster communication system, enabling all the populations in an affected area to signal their safety status, provide vital intelligence on the disaster, and request emergency assistance without saturating emergency call centers (PSAPs).
EU-ALERT is a standardize system capable to broadcast an emergency message to all the smartphones of a defined area in case of major disaster. It has been as a way for the authorities to communicate toward the population (top-down) but never receive answers from them, having to rely on standard 112 emergency calls that get saturated during disasters.
To empower the population during a crisis, the GDPR-compliant module is integrated in the EU-ALERT messages through a web link (URL) allowing the population to signal their safety status (safe, seeing a danger, in an emergency), share their location, provide pictures of the situation and call emergency services if required. HighWind patented computer vision’ artificial intelligence analyze all the pictures that are collected to provide a pre-diagnostic estimating the nature & severity of each situation based on a cross detection of Traumas, Context and Emotions. The AI support emergency triage and decision-making of the crisis management cell, highlighting most vulnerable persons and situations to be dealt with. With consecutive reporting, Advance-ALERT allows to track the movements of the crisis’ epicenter (ex: how a wild fire progress) leveraging on the presence of the population on the ground.
This system has been already tested in France during FR-ALERT exercices, being broadcasted to over 300 000 persons, with exceptional and spontaneous participation of all the population in the targeted area. With EU aiming to create a true Disaster Resilient society, Advance-ALERT is an essential tool, that can be integrated in matter of seconds into the existing systems, and leverage on all the population to save lives.
Advance-ALERT is an innovative solution addressing the limitations of the official and standardized European disaster communication system EU-ALERT, that until now only allowed top-down communication toward the population. The project ambition is in line with creating a European Disaster Resilient Society, as it empowers the population to provide critical intelligence on the ground to the authorities/emergency services, avoid jamming the emergency call centers, and maximize the efficiency of life-saving evacuations.
The solution is designed to address the emergency communication needs of disaster relief efforts while at the same time complying with the highest standards in terms of European data protection (GDPR). Tested several times since prototyping, with the support of French local authorities, Advance-ALERT received the immediate and spontaneous commitment of the population of the defined areas, understanding immediately the utmost importance of communicating back with the authorities.
Advance-ALERT represent a shift in the state of mind of disaster relief efforts, the population is no longer considered to be “victims” waiting to be rescued, but instead makes them key actors empowered with enhancing the relief efforts and providing vital information for the first-responders. The project emphasize the simplicity of deployment and utilization, from both the population and the authorities. Both interfaces (population and authorities’ digital map) are web-based, can be integrated in EU-ALERT just by copy/pasting a web link and are immediately ready with standing capacity to cope with 500k almost simultaneous disaster signal. Authorities can share their global view with all required services with a mere link and dual secured authentication and AI is the power-house of the system.
The powerhouse of the solution is a patented artificial intelligence specialized in the pre-diagnostics of emergency situation through computer vision detection in pictures provided by the population. The artificial intelligence analyse the situation under three axis : (1) TRAUMAS : all external wounds present on persons, like bleeding, cuts, dislocations, burns, etc. ; (2) CONTEXT: all the elements surround the emergency, like flames, smoke, mud, water, corpses on the ground, paralysis, etc. ; (3) EMOTIONS : pain level, fear, stress, etc. By cross-analyzing these three axis, the AI is capable to provide a nature and severity of the emergency situation of each person signaling herself to the emergency services, allowing the emergency services to confirm critical assessed level of each reported situation, quickly determine most adequate team to intervene, grasp side-effects of the crisis.
Faced with possible hundreds of thousands of simultaneous emergency signals, in limited areas, the AI is the mandatory tool to give meaning to the retrieved information instantly. So far the AI has demonstrated being able to assess a situation in <90ms, when the most efficient emergency calls last 2.5min and can last for up to 10min, and while being carried by the EU-ALERT systems and sized for large date, the system has demonstrated during large-scale exercises to be able to retrieve several thousand of signals in 3 minutes.
As demonstrated during the large scale exercices led by French authorities, Advance-ALERT has proven :
Large scale diffusion/reception : being carried by FR-ALERT disaster communication system, the solution was able to address all the population located within the Alpes de Hautes Provence’ department (200k inhabitants) in matter of seconds, while being able to retrieve signals from the population and display it in less than 2 minutes
quick set-up: being fully cloud based and having its population and authorities interfaces that can be opened through dedicated web URLs, Advance-ALERT demonstrated during the Elengy Exercice (Bouches-du-Rhône, November 21st 2024) operational readiness in less than 5minutes between the support request by the authorities to HighWind and the broadcast of the FR-ALERT message including Advance-ALERT module.
high resilience : advance-ALERT has been designed to cope with the chaotic environment of disaster situations, including redundancies of information collection, capacity to operate with scarce network availability (degrading image quality while maintaining AI analysis capacities), adding a new layer to the existing emergency communications
Advance-ALERT address the dire need for the population to communicate their situation to the emergency services and authorities that must organize disaster relief efforts, while avoiding saturation of networks and emergency call centers. The solution was already tested on various scenarios including natural disasters and industrial accidents, and will become an essential asset to coordinate rescue efforts during future crisis. Instead of being merely victims, the population is empowered with providing critical information to the authorities having to coordinated the disaster relief efforts, giving thousands of eyes on the ground that are then converted into vital information through the AI.
The team encompass AI and disaster management experts from HighWind and legal data protection experts from GAGDPR. This combination is vital for the solution as it is massively broadcasted to the population, retrieves sensitive information, analyze it by AI and provide support to emergency triage, with possible massive legal consequences. The team worked therefore closely to design a solution fully compliant with EU regulations (Electronic Code of Communication, GDPR, AI Act, etc.).
The team encompass AI and disaster management experts from HighWind and legal data protection experts from GAGDPR. This combination is vital for the solution as it is massively broadcasted to the population, retrieves sensitive information, analyze it by AI and provide support to emergency triage, with possible massive legal consequences. The team worked therefore closely to design a solution fully compliant with EU regulations (Electronic Code of Communication, GDPR, AI Act, etc.).