**2. Why natural disasters management must be supported by VR/AR**

Built environments, which refer to all physical environments constructed for human habitation and activities [9], are constantly exposed to risk from various natural and manmade disasters, such as fires, tropical cyclones (wind and storm surge), earthquakes, tsunamis, floods, and terrorist attacks, which pose a significant threat to human beings. Considering the severity of natural hazards, that become disasters when people's lives and livelihoods are destroyed, an appropriate NDM is crucial to reduce the harmful effect of the phenomena and to facilitate a prompt reestablishment of the normal life after the emergencies [10]. When the threatened scenarios are large cities or strategic sites for nations, it is essential to have efficient and timely management programs that protect lives and properties of the inhabitants but also that quickly restore the productive and commercial capacities of population centers [11]. NDM covers i) risk prevention, ii) crisis preparedness (specific training), iii) emergency response (rescue), and iv) catastrophe recovery (reinstatement of services and lifelines) [12]. The key point in these actions is the preparation facing of the emergency so the researchers and professionals must work to improve the understanding of the situations, processes behind and analysis methods, anticipating the adverse possible situations, their worst evolution, and forecasting how they arrive to critical states.

The NDM process to save human lives, divided in pre-, during-, and post-disaster actions, still have a gap between knowledge and *intelligent* and effective procedures [13]. The impossibility to prepare and train under real-life conditions, limits the learning experience of first-responders, civilians, and city-planners. During

#### *Intelligent VR-AR for Natural Disasters Management DOI: http://dx.doi.org/10.5772/intechopen.99337*

disasters, communication, and visualization of possible pathways to recovery (since danger could be permanent and/or resources are scattered) is fundamental. In this sense, VR and AR technologies cannot be ignored anymore: they have grown exponentially over multiple markets and is projected to have a worldwide revenue of USD 40 billion by 2024 [14, 15]. The examples of successful applications are varied and motivating: simulation of floods, ground motions due to earthquakes, wildfires and hurricanes, around the training, the monitoring, the modeling and the early warning [16–28]. The virtual and augmented scenarios, most of the times, are developed as smart systems (based on artificial intelligence) to improve situations prediction [29–34].

Between the various studies the use of smartphones is indispensable. The potential and benefits of using AR to access and display disaster data in its geospatial context [35–37] is normally developed through a mobile application, for example to determine the best route for evacuation from collapsed or burned areas. The Whistland system [38] is an example of retrieving crowdsourced disasterrelated information from real-time Twitter feed for display as an AR overlay to the smartphone camera. Mirauda et al. [39] developed a mobile application for connecting measurements from hydrological sensors with an integrated forecasting model. Fedorov et al. [40] presented a unique approach to utilize computer vision techniques to identify mountain silhouettes and compare the extracted information to available DEM (digital elevation model) data to provide useful information (e.g., peak name, height, lithology, associated risks, etc.).

About AR/VR solutions the best models are those developed for visualizing data and analyzing information in-situ in smooth transitions between virtual and augmented environments [41]. Ready et al. [42] presented a virtual reality application for HTC Vive that recreates a 3D model of terrain and buildings (geographic context in Japan) to interact with various data resources, mainly to access hydrological time-series in an easily interpretable way for disaster management. Haynes et al. [43] developed a mobile application to visualize potential floods through integration of real-time recordings (e.g., water level, soil moisture, and humidity), being extremely important that stakeholders can observe situations from the application and make decisions consistent with what happens in the field. Macchione et al. [44] proposed a virtual environment (open-source 3D graphics app, i.e., Blender) to recreate an urban environment (e.g., buildings, streams, roads, levees, textures) to simulate hydraulic dynamics during different flood scenarios.
