In fleet management, fleet maintenance is a crucial aspect to increase availability. Periodic and preventive

Many businesses are becoming increasingly interested in managing the supply chain to survive in today's dynamic market. Of the many factors that influence supply network maintenance, demand forecasting forms the origin of all activities in the supply chain. It has, however, been irregular and inadequate at forecasting in a dynamic market. The intermittent and irregular characteristics of replacement demand for products and parts make forecasting demand particularly difficult. Moreover, as information concerning demand from a specific period is often required on site, a demand forecasting method unlike those available at present is needed.In this study, an Internet of Things environment is built and a technique proposed to forecast replacement demand in a specific period featuring intermittent and irregular demand characteristics. The study measured the lifespans of batteries, and compared the Weibull distribution, artificial neural networks (ANNs), and recurrent artificial neural networks.The ANN had a shorter application period and higher forecast rate than the demand forecasting method employing the Weibull distribution. Further, when this network was circulated, the forecast rate improved over the general ANN.

With the recent increasing demand of freight, parcel delivery service, and product distribution, the quantity of supply that a company needs to handle is also increasing. The travel time between the demand points in the city is greatly affected by the complicated road conditions and the changing external environment in real time. In the previous researches, the vehicle routing problem is to define a route based on the distance between the demand points, the travel time, and the determined demand. This study propose a vehicle routing problem considering the vehicle location, vehicle condition information and dynamic demand forecasting by the Internet of Things (IoT) equipment attached to the vehicle.