Machine Learning Approach for Charging Queue Waiting Time Prediction of Electrical Autonomous Forklifts Fleet

Abstract

The availability of electrical autonomous mobile robots is a main topic addressed by many researchers. Notably, the battery-powered Forklifts, recognized as battery consumers, have disadvantages of short autonomy and long charging time, which would require an accurate prediction of the charging queue waiting time to optimize the charging strategy. This paper proposes a machine learning approach to predict the charging queue waiting time for the electrical autonomous Forklifts (EAFLs) fleet. Results show accurate predictions with a mean absolute error of 1.52 min in comparison to the naive mean and median methods, which achieved a mean absolute error of, respectively, 5.54 min and 5.46 min.