SUMMARY

Nowadays, E-commerce is vastly popularized, and people can easily shop online from anywhere in the world. The progress of globalization also makes most commodities contain parts that are made from different places of the world. Freight transport by truck plays a vital role in such activities. Increasing the efficiency and reducing the waste of such transportation has a significant impact on not only increasing the profit of the transportation company but also making it more friendly to the environment.

To increase such efficiency, this thesis proposed a Smart Fleet Dispatching and Route Planning System. This is established with two aspects: The Real-Time Data Collection and Processing and The Fuel Consumption Rate Focused Route Planning Model. The first aspects enable the truck to collect real-time data such as weight and traffic conditions. The weight is attained by frequency domain analysis, and the traffic condition is attained by neural networks. The second aspect uses the data from the first aspect to create a model that is better represented in practice than the conventional route planning model. As the cost of transportation is majorly determined by the fuel consumed, and the fuel consumption rate can be assumed to be linearly related to the weight and the traffic condition. The thesis verified the performance of the proposed model via simulation, which compares the fuel used with the output plan from the conventional Capacitated Vehicle Route Planning model.