Analisis Pemetaan dan Prediksi Human Error Dari Pengemudi Sepeda Motor terhadap Kecelakaan Lalu Lintas Menggunakan Bayesian Networks (Studi Kasus: Kota Medan)

Abstract

Statistical data shows that 50 million people are injured and 1.35 million people die in traffic accidents every year in the world. Indonesia is the country with the fifth highest number of traffic accidents in the world. The Indonesian National Police (2021) states that most traffic accidents involve motorcycle riders. Traffic accidents in Medan City by motorcyclists are caused by several factors including human or driver factors (93.52%), vehicle factors (2.76%), road factors (3.23%), and environmental or weather factors. (0.49%). Medan City Traffic Traffic Unit has implemented five traffic accident prevention programs, namely mapping areas with potential traffic accidents, Global Road Safety campaign, establishing traffic forums, revitalizing traffic orderly areas, and Road Safety Partnership Action. However, these five efforts have not been able to significantly reduce the number of traffic accidents. The high number of traffic accidents in the city of Medan is mostly caused by the human factor or the driver. Therefore, research on motorcycle driver factors (human error) needs to be done to be able to obtain a prediction of future traffic accidents. This research focuses on human errors from motorbike drivers which can be divided into five, namely traffic errors, control errors, speed violations, safety violations, and traffic violations. Bayesian Networks (BNs) are used to predict the number of traffic accidents during 2023 due to human error. Based on the results of calculations using Bayesian Networks, it is predicted that the most traffic accidents in 2023 involve motorcycles with traffic errors (51%) and traffic violations (21%). The category of traffic accidents caused by being careless when turning (18%) and driving without using an SNI helmet (18%). To reduce the number of traffic accidents and the number of drivers injured each year, the recommendations given are to improve the quality of the Global Road Safety campaign, revitalize traffic orderly areas, and make use of prediction templates to predict the number of traffic accidents in the future.