In this manuscript we propose a warning system based on Machine Learning (ML) techniques to allow early detection of volcanic eruptions. The use of Artificial Intelligence (AI)-based algorithms and the early detection purpose can be exploited together to carry out the specific detection and inform the Civil Protection and the authorities in charge to guarantee immediate intervention for the protection of people. Indeed, the main goal presented in this work is to conceptualize and realize a system to be mounted on-board of satellites with the intent of producing swift alerts useful for decision makers. In the initial phase the proposed ML model is trained with SO2 data acquired by Sentine1-5P. The final scope is to define an on-board system similar to Phi-Sat 1, with its own sensor and its own systems. Our feasibility study has shown that it is possible to classify and identify volcanic eruptions in advance with an accuracy of 80% and 70% respectively. Work is in progress to improve these results.

Early Detection of Volcanic Eruption through Artificial Intelligence on board

Sebastianelli A.
Membro del Collaboration Group
;
Ullo S. L.
Supervision
2022-01-01

Abstract

In this manuscript we propose a warning system based on Machine Learning (ML) techniques to allow early detection of volcanic eruptions. The use of Artificial Intelligence (AI)-based algorithms and the early detection purpose can be exploited together to carry out the specific detection and inform the Civil Protection and the authorities in charge to guarantee immediate intervention for the protection of people. Indeed, the main goal presented in this work is to conceptualize and realize a system to be mounted on-board of satellites with the intent of producing swift alerts useful for decision makers. In the initial phase the proposed ML model is trained with SO2 data acquired by Sentine1-5P. The final scope is to define an on-board system similar to Phi-Sat 1, with its own sensor and its own systems. Our feasibility study has shown that it is possible to classify and identify volcanic eruptions in advance with an accuracy of 80% and 70% respectively. Work is in progress to improve these results.
2022
978-1-6654-8574-6
Artificial Intelligence on board
Early Detection
Machine Learning
Phi-Sat 1
Sentine1-5P
File in questo prodotto:
Non ci sono file associati a questo prodotto.

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/20.500.12070/57560
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 0
  • ???jsp.display-item.citation.isi??? 0
social impact