首页|Using near misses, artificial intelligence, and machine learning to predict maritime incidents: A U.S. Coast Guard case study
Using near misses, artificial intelligence, and machine learning to predict maritime incidents: A U.S. Coast Guard case study
扫码查看
点击上方二维码区域，可以放大扫码查看
原文链接
NETL
NSTL
Wiley
Two recent trends made this project possible: (1) The recognition that near misses canbe predictors of future negative events and (2) enhanced artificial intelligence (AI) andmachine learning (ML) tools that make data analytics accessible for many organizations.Increasingly, organizations are learning from prior incidents to improve safetyand reduce accidents. The U.S. Coast Guard (USCG) uses a reporting system calledthe Marine Information for Safety and Law Enforcement (MISLE) database. Becausemany of the incidents that appear in this database are minor ones, this project initiallyfocused on determining if near misses in MISLE could be predictors of future accidents.The analysis showed that recent near-miss counts are useful for predicting futureserious casualties at the waterway level. Using this finding, a predictive AI/ML modelwas built for each waterway type by vessel combination. Random forest decision treeAI/ML models were used to identify waterways at significant accident risk. An R-basedpredictive model was designed to be run monthly, using data from prior months to makefuture predictions. The prediction models were trained on data from 2007 to 2022 andtested on 10 months of data from 2022, where prior months were added to test the nextmonth. The overall accuracy of the predictions was 92%–99.9%, depending on modelcharacteristics. The predictions of the models were considered accurate enough to bepotentially useful in future prevention efforts for the USCG and may be generalizableto other industries that have near-miss data and a desire to identify and manage risks.
NETL
NSTL
Wiley
