Case Study

The project

We were tasked with helping a large appliance manufacturer to optimize their enameling process so that it uses less energy while ensuring the quality of the enamel. By using our Digital Twin platform, our team modeled the Conveyor, Combustor, and firing Zone. All of the heat transfer processes were also modeled -  convective transfer between the coils and air, convective transfer between the air and walls of the furnace, convective transfer between the air and load, radiative transfer between the coils and walls, radiative transfer between the coils and load, radiative transfer between the walls and load, radiative transfer between flue pipes and load, thermal mass flow between the combustor and coils, thermal mass flow between the coils and stack. This provided us with a very accurate simulated environment that we could use to train our AI agents to optimize the process. 

​

The algorithms found that by reduction of the set point of two of the firing zones or the reduction of the fuel-air ratio (through the adjustment of the valves that control the air and gas supply) leads to fuel savings of up to 22.37% and 7.2% and electricity savings of 22.38% and 2% respectively.