Model for improving the logistics processes for propane delivery

The scheduling problem of service times and the method to provide service to customers by minimizing time, cost, and distance are important for service providers and are challenging research topics. This research problem concerns the scheduling of the delivery of propane to a large number of customers in a rural area. Prediction of customer demands must be done before service times are scheduled. In this research, there are two types of customer demands; one is regular customer demands and the other is call-in customer demands. This problem is similar to a situation having regular tasks and emergency tasks that come up to interrupt the regular tasks. The regular customer demands are scheduled and the regular service is assumed to be the routine every day. The data collection of regular customer demands are observed from historical demands of regular customers and separated by season as summer and winter. Afterwards, the proper mean (ë) number of call-in customers per day is calculated, and it is established that the Poisson distribution applies. Subsequently the call-in services are scheduled after these customers call for service. The challenge of this research is that each customer service time is scheduled in the time limit per day. In addition, the locations of regular customers are investigated before grouping regular customers together to provide service on the same day. This research focuses on the method to integrate the schedule containing regular customers with the call-in customers in each day.