Forecasting Transportation Volumes at Farm Trans

The main research question of this research is: How can Farm Trans predict the transportation volumes within the conditioned business unit with the available information? Forecasts are generated on multiple aggregation levels, corresponding to the decision they will support. This is to support fleet capacity decision-making. Forecasting models and methods like ARIMA, random forest and hierarchical forecasting methods are assessed to find the best according to the accuracy of the test data. Furthermore, a prototype for implementation is developed.