Study goals
The objective is to verify the association between machine learning and a wood road transport database in predicting the net mass of wood transported by vehicular load combinations.
Relevance / originality
Machine learning algorithms are widely used to solve road transport barriers, however, they are rarely applied in the prediction of transported net mass.
Methodology / approach
The input variables were composed by age, time after cutting, and wood density. To estimate the net mass of transported wood, random forest machine learning algorithms, support vector machine, and K Nearest Neighbor were used.
Main results
The support vector machine algorithm performed best in the prediction models of the net mass of wood transported by vehicular load combinations on highways.
Theoretical / methodological contributions
The association between machine learning algorithms with a wood road transport control database promoted the prediction of the net mass of wood transported by vehicular load combinations.
Social / management contributions
The proposed technology promotes increased precision in the operation of road transport of wood. It allows management to act more efficiently and standardize its processes, reducing possible penalties for employees due to the automation of the activity.