REFINING PROCESS MINING IN PORT CONTAINER TERMINALS THROUGH CLARIFICATION OF ACTIVITY BOUNDARIES WITH DOUBLE-POINT TIMESTAMPS

Port container terminals play a vital role in global trade, necessitating effective analysis techniques to optimize operational processes. Process mining methodologies have been widely employed for this purpose, but most existing approaches rely on singlepoint timestamps, lacking detailed information on activity durations. To overcome this limitation, this paper proposes a new methodology that utilizes the PERT-Beta Distribution to determine a double-point timestamp event log from an organization’s traditional database. By incorporating double-point timestamps, the proposed methodology provides a more comprehensive understanding of activity durations. The PERT-Beta Distribution calculates the mean and variance of time for each activity by considering optimistic, most likely, and pessimistic time estimates, along with shape parameters alpha and beta. This approach accounts for uncertainty and enables effective management of project completion time. Additionally, the paper presents a process model discovered using an algorithm based on the double-point timestamp approach. To validate the proposed methodology, real-world data from Terminal Operating Systems (TOS) in Surabaya, Indonesia, are utilized. This research contributes to the field by enhancing process mining techniques through the introduction of a novel methodology that incorporates double-point timestamps and leverages the PERT-Beta Distribution. The case study conducted on a port container terminal provides practical insights for improving operational processes and efficiently managing project completion time.