This study presents our most recent advances in the design of a data-driven method for mass flow rate estimation of conveyor belts. Our proposal is focused on obtaining an indirect method that uses power measurement from the conveyor belt. The aim is to replace traditional expensive measurement hardware, which results in benefits such as lowering overall costs as well as the possibility of working in hostile environments such those with adverse weather conditions and the presence of dust and vibration. The mass estimation is based on data-driven estimations of idle power and net energy consumption. We discuss different models describing the relationship between energy input and transported mass: a constant proportionality factor, a time-dependent factor and a regression model depending on the idle power. We illustrate our approach on a case study where the state-dependent model yields the most promising results across multiple working periods.