Optimization of the use of support vehicles based on stochastic demand forecasts (#739)

Authors

De Santis Bermeo, David Antonio

Calligana Chancay, Roxanna

Darío, Tigrero Carpio

Abstract

This paper seeks to propose a methodology that allows determining the optimal number of vehicles to hire, considering the following assumptions: stochastic demand, flexibility for vehicle contracting, homogeneous fleet of vehicles The methodology is based on an analysis of qualitative. and quantitative variables that represent the demand such as the number of clients to visit and the demand of each of these clients. We proposed the analysis using two approaches, one that took into account the days of the week and another that considers the date in calendar days. Based on this, a probability distribution model was found, which served us for the purification of the data. The selection of the best time series model was made, based on the predicted number of clients to visit. The clients were classified according to their demand and together with the forecast of the clients to visit, the number of vehicles that will be needed for each period was estimated. All this methodology is developed through a case study, where the savings generated are saw, comparing with the current situation.