A novel fuel savings training needs prediction and alert system is capable of calculating a probability of persistent future fuel waste of a particular driver based on recent driving patterns originating from real-time on-board diagnostics (OBD) information of a vehicle. The fuel savings training needs prediction and alert system is also configured to generate and transmit one or more action items and alerts to a vehicle fleet operations manager or another supervisory personnel for improving driving behaviors of the particular driver. The fuel savings training needs prediction and alert system is typically either linked to or incorporated into the fuel waste variable identification and analysis system, which is capable of analyzing the real-time on-board diagnostics (OBD) information of the vehicle from a remote monitoring station unit. Preferably, various fuel waste variables are normalized with a novel metric called “RPG,” or “engine rotation per minute per gallon of fuel.”