Assessing the Risk of Crashes on GCC Roads with a Markov-Based Method

Road traffic accidents represent a serious problem all over the world, causing several losses in many ways. The situation in United Arab Emirates (UAE) is not an exception. Therefore, this paper was developed providing a frequencies analysis for accidents data within five successive years on Abu Dhabi roads. Markov modeling was used to find the transition probability from either the (week, month, or year) to the next level, and the steady state probability was estimated using a long run probability. Two policies were used to reduce the number of accidents. It was concluded that the sojourn time for the first policy was better than the second one. Therefore, for the monthly and yearly transition probability matrix only the first policy was applied. Further research should consider the total vehicle number to find the accident rates and identify the significant reasons of accidents to determine the suitable actions.