Assessing Offshore Oil Spills Using Remote Sensing and Geospatial Artificial Intelligence (GeoAI): A Systematic Literature Review

Alva A.I.; Rodríguez C. · Ingenierie des Systemes d'Information (2024) · DOI: 10.18280/isi.290602

Resumen

Artificial Intelligence (AI) cuts across all disciplines of human knowledge, which favors the development of new approaches to the care and preservation of the environment. Oil spills at sea, caused by the transportation and marketing of hydrocarbons, have been considered an unstoppable source of pollution in the last century. So, the present Systematic Review of the Literature (SRL) analyses the most recent publications in Remote Sensing (RS) and AI (Machine Learning, ML) and Deep Learning, DL) identifying a total of 3,355 published articles, and classified 50 articles based on new taxonomy proposed in the study. In this way, it identifies two main Offshore Oil Spill Datasets (OOSD): the HOSD and MKLab projects. Also, it classifies the literature that uses ML at 14%, and DL at 46%, which implies 40% for traditional RS at Offshore Oil Spills (OOS) Detection. This leads RS & AI to consolidate under the concept of Geospatial Artificial Intelligence (GeoAI), a new field of research that allows, in addition to detecting and predicting offshore oil spills under the best performance metrics, such as Precision, Machine Learning 96.5 %, and Deep Learning (99.19 %) using U-Net Model.

Instituciones: Universidad Nacional Mayor de San Marcos