Flood events management in alpine areas by means of innovative methodologies using IoT

Natural or man-made disasters can be lethal for environment, infrastructures and even human life. The proposal of this research is to analyse the repetitive flood events in Cervo Valley, an alpine Area in the North-west of Italy, in order to determine a methodology able to reduce economic and social damages. Floods events have always existed and their negative consequences for the society are related to the interference between human activity and natural river flow. In the recent past, the strategy for limitation of flood associated damages was based on the protection of areas of interest by means of structural measures that physically prevented water from flowing into them. Climate changes of the last decades and latest studies on this subject highlighted the partial ineffectiveness of this strategy. For this reason, the approach to natural hazard must shift from a defensive action to management of the risk and live with it. This can be achieved favouring prevention and mitigation by establishing an automatic procedure allowing to get real-time field inputs and send back a response on how to manage the emergency. In fact, data shows that management measurements tend to be more efficient and produce more long-term and sustainable solutions. Technologically speaking, the recent advancements of data acquisition sensors, the introduction of satellites data, the improvements of BIM and GIS methodologies, the Internet of Things (IoT) and the evolution of big data analysis technologies (BDA) and machine learning (ML) have provided an open opportunity to innovate data harvesting, aggregation and processing procedures. Furthermore, the combination of these technologies will allow to create new tools for emergency response. In fact, nowadays information coming from heterogeneous sources are easily collectable in real-time. If this huge volume of data is properly managed, it allows to respond promptly to an extreme event. First of all, it is necessary to evaluate the risk assessment of the considered region. That allows to be acquainted with the valley and recognize the more vulnerable zones. On the basis of the area evaluation it is then possible to decide which are the most suitable recording instruments to be employed and their amount, remembering that data collection is fundamental for a reasonable system response. IoT allows to create a real-time database that can be directly associated with a digital twin of the valley, where all the general information is stocked. The digital twin is created taking advantage of the interoperability between BIM and GIS systems. Finally, it is necessary to analyse collected data in order to fit them in a pre-defined function that permits to find the most suitable scenario. In that sense, BDA and ML must be implemented to obtain a prompt response and to improve it step by step. To summarize, this research shows a methodological framework implementing the abovementioned technologies, which has been applied to the case study of Cervo Valley. From such application, the proposed methodology resulted to be replicable in other alpine valleys varying some of the parameters according to the characteristics of each valley.