The Autoregressive Integrated Moving Average (ARIMA) model as a tool for predicting accidents in the maritime domain: the case of the Galician fishing fleet

Abstract

[Abstract] The Autoregressive Integrated Moving Average (ARIMA) model has proven to be a powerful statisti-cal prediction technique due to its simplicity and wide acceptance. Its main application lies within therealms of economics and social sciences, although not limited to these. This study explores its utiliza-tion in the context of maritime safety within the Galician fishing fleet. Specifically, its application isproposed for predicting the vessels involved in accidents quarterly index.The Galician fishing environment, of significant socioeconomic relevance to the region, is affected byserious accidents compromising its sustainability. The vessels involved in accidents index, derivedfrom investigation reports by the Maritime Accident Investigation Commission (CIAIM) and fleet datacollected by the European Fishing Vessel Register, depicts the annual evolution of these accidents inrelation to the decreasing fishing capacity of the fleet. Analysis of quarterly values for the periodbetween 2011 and 2021 indicates that accidents are recurrent and persistent in nature.Using the proposed ARIMA model, index values for the quarters of 2022 were forecasted and themodel’s effectiveness was validated by comparing the obtained results with actual observed data.In the studied case, this predictive capacity supports proactive safety management on board while pro-viding a solid foundation for the adoption of preventive initiatives in both public and private sectorswithin the industry.