Cyber-Physical Systems (CPS) are defined as the integration of computational and physical processes, whereas the concept of using a digital copy of a physical system to perform real-time optimization is referred to as a  Digital Twin. 

Digital Twin design and development requires Mathematical model of a given physical process.  Neural Network is getting used to model a system by using a given data set and associated labels on a given data set.  Pontryagin duality and Pontryagin Observer appear to be helpful in creating Digital Twin.  

Boltzmann machine is used to model a physical process by using a Neural Network.  Whether it be probability, statistics, Data Science, Machine Learning, Deep Learning, or any other likewise field, having the knowledge of the distribution of data is a must or crucial, because it helps in dealing with data.