Ensuring Excellence in Natural Disaster Management (NDM) in Europe

NDM-pioneers in Europe 

In Europe, the occurrence of extreme events, like forest fires and floods, is ever increasing in magnitude and damage severity. While contemporary technologies appear matured, new tools must be developed to more accurately, efficiently and effectively manage ND. TEMA aims to achieve this at European level: with its pan-European, fully qualified and multidisciplinary consortium, TEMA will develop an NDM-AaaS functionality that will facilitate NDM platform interoperability across European countries and regions.

Toward excellence in NDM

NDM can be greatly improved by developing automated means for precise semantic mapping and phenomenon evolution predictions in real-time. Several extreme data sources can significantly help towards achieving this goal: autonomous devices and smart sensors at the edge, equipped with AI capabilities; satellite images; topographical data; official meteorological data; predictions or warnings published in the Web; and geosocial media data. Such heterogeneous data sources provide a prime extreme data example: 

• they are diverse (images, videos, text, scientific measurements); 
• voluminous (images and videos), 
• fast (most of them come in real-time) and frequently updated;
• complex (many of them are unstructured);
• multilingual (text);
• have very disperse sources (satellite, drone, sensors, social media, maps);
• are sparse/missing (as forest fires and floods can only be sparsely observed in space and time); and
• have extreme values (by definition, as forest fires and floods are exceptional events).

TEMA aims to develop state-of-the-art technologies facilitating this vision, focusing on real-time semantic extraction from multiple heterogeneous data modalities and sources, on-the-fly construction of a meaningful semantically annotated area map, near-real-time prediction of phenomenon/emergency evolution and automated response recommendations. Semantic analysis computations will be distributed across the edge-to-cloud continuum according to emergency response needs, in a federated manner to keep latency to a minimum. Heterogeneous data analytics will be performed, mostly based on artificial intelligence (AI) and Deep Neural Networks (DNNs) in a trustworthy, transparent and flexible way, deployed as fit to the emergency and catering to various user needs. The constantly updated 3D map, the predictions and the recommendations will be used as the basis for an advanced, interactive, Extended Reality (XR)-based interface, where the current situation will be visualized and different alternatives for device/sensor deployment and response strategies can be dynamically evaluated by human operators. The end result will be a SoA NDM support system, dynamically exploiting multiple data sources and AI technologies for providing an accurate assessment of an evolving crisis situation.

NDM models, data analysis and response require compute-intensive machines to support and run the ensemble simulations and aggregate the results for better interpretation. Fast computations are crucial TEMA system elements, as their outcome supports decision making on evacuation and action plans. Cloud computing, with its almost unlimited capacity of computation, storage, and networking resources to handle the associated challenges, provides new NDM capabilities (Ujjwal, 2019). In TEMA, the entire edge-to-cloud continuum will be spanned, to meet the extreme data analytics and response needs. The overall cloud and edge computing infrastructure can offer scalability, performance, storage, ubiquitous access and security.

TEMA will provide heads-up, hands-free, contextualized operational support in a wide range of environments to fasten and ease the realization of complex tasks while making them safer. A highly advanced and versatile Augmented Reality (AR) interface will be developed for integrating and visualizing live all TEMA results (semantically annotated 3D map, predicted outcomes, AI prediction explanations and proposed recommendations), as support for the human user in an operational “control room”, while also allowing them to interactively assess contingent response alternatives via simulation. TEMA will deliver a technical solution to make modern AR systems exploitable in disaster response and management, bringing situational data to relevant end-users, thus providing the relevant information that can help make the best possible operative decisions.            

Key TEMA-characteristics:

• TEMA will be an AI-enabled NDM platform which will be validated and demonstrated in two socially important cases: forest fires and floods, in four locations spread all over Europe (Germany, Finland, Italy and Greece);
• Due to the scale and diversity of the employed extreme data and their powerful analysis, TEMA will offer precise semantic 3D area mapping and NDM event prediction over a longer time period (before/during the event), by taking into account all related uncertainty factors (e.g., meteorological ones) TEMA tools will allow precise replication and modeling of real NDM events, while considering quality standards and assessment criteria for simulated NDM data;
• Advanced AR-enabled TEMA visualization will be interactive, intuitive and accessible, enabling human operators to understand complex NDM phenomena and easily evaluate alternative response strategies, by smart selection of platform parameters and by novel ways of combining heterogeneous (visual and non-visual) data and/or AR;
• TEMA will provide a scalable and efficient NDM-Analytics-as-a-Service (NDM-AaaS) for making the developed advanced analytics tools available over the cloud.

Take-homes:

The problem: TEMA addresses NDM-needs by developing automated means for precise semantic area mapping and phenomenon evolution predictions for NDM in (near-)real-time;

The need to address NDM at European Level: The disaster events to be handled by TEMA happen all over Europe. Therefore, research can be best done at European level; this, together with the foreseen TEMA NDM-AaaS functionality, will greatly facilitate NDM platform interoperability across EU countries and regions;

The time is now: The occurrence of extreme events facing the climate trends is ever increasing in magnitude and damage severity; this can be matched by the timely development of mature AI, XAI, XR technologies to handle extreme data analytics for solving NDM problems;

Technological availability and preparedness: Data analytics, AI and AR/XR tools are mature; albeit, the degree of extreme data analytics automation cannot be realized with existing tools; new research must be done and novel NDM tools be developed;

Why TEMA: Due to the groundbreaking TEMA objectives, only a pan-European, perfectly balanced, fully qualified and multidisciplinary consortium, composed of experts in all relevant specific subfields (data analytics, AI/machine learning, remote sensing/Earth Observation, federated and cloud computing, fire/flood modeling, geovisual analytics, AR, NDM), would be able to properly face the challenge. They come from leading research laboratories with significant international R&D experience and relevant end-users from across Europe.