Large consumers including factories use various utilities such as energy (electricity, heat, fuel, etc.), water and compressed air, and they need to not only reduce energy costs, but to also simultaneously solve many difficult issues including the reduction of CO2 emissions, the improvement of production efficiency, the stabilization of operations and the shortage of skilled engineers. To solve these issues, prediction technologies for determining how much and when energy will be needed in the future are essential. Mitsubishi Heavy Industries Ltd. (MHI) developed the "power demand prediction system" which predicts factory power demand using its proprietary AI, and this report introduces the system as well as case examples of demonstrations conducted in MHI's plant.