3D digital twin allows industrial user to visualize in 3D and real-time all areas of production using our virtual sensor capability, including areas of difficult access and areas without physical sensors, so that the onset of any production problem can be easily determined by virtually zooming to any component of the machine to understand the problem root cause and run a couple of simulations to solve it. The virtual sensor technology uses a 3D physics-based reinforcement learning calculation system embedded in the software to determine the values of the engineering parameters in the areas without sensors and it can be used in real-time to improve the accuracy of the control algorithm by adding much more data points (virtual sensors) on top of physical sensors, which allows improving the energy efficiency of the industrial production process between 20% to 30%.