Training and refining fluid models using disparate and aggregated machine data

A method, comprising:

• generating, by a system comprising a processor, a three-dimensional model of a mechanical device based on a library of stored data elements, wherein the data elements each comprise one or more properties of the mechanical device, and wherein generating a three-dimensional model further comprises determining a set of one or more control volumes associated with the mechanical device, the set of one or more control volumes abstracting a region of the mechanical device through which a fluid and/or an electrical current flows;
• predicting, by the system, fluid flow and physics behavior associated with the three-dimensional model based on a machine learning process associated with the three-dimensional model;
• rendering, by the system, physics modeling data of the mechanical device within a three-dimensional design environment based on the fluid flow and the physics behavior;
• receiving, by the system, machine data from each of a plurality of additional mechanical devices, wherein the machine data is received via a network device of a communication network;
• and
• updating, by the system, the physics modeling data and the set of one or more control volumes within the three-dimensional design environment based on an aggregation of the machine data collected from the plurality of additional mechanical devices, wherein the plurality of additional mechanical devices and the mechanical device are each a same type of mechanical device.

A non-transitory computer readable device comprising instructions that, in response to execution, cause a system comprising a processor to perform operations, comprising:

• generating a three-dimensional model of a mechanical device based on a library of stored data elements, wherein the data elements each comprise one or more properties of the mechanical device, and wherein generating a three-dimensional model further comprises determining a set of one or more control volumes associated with the mechanical device, the set of one or more control volumes abstracting a region of the mechanical device through which a fluid and/or an electrical current flows;
• performing a first machine learning process associated with the three-dimensional model to predict one or more characteristics of the mechanical device;
• generating physics modeling data for the mechanical device based on the first machine learning process;
• receiving, from a network device of a communication network, machine data from each of a plurality of additional mechanical devices;
• updating the physics modeling data to generate updated physics modeling data and the set of one or more control for the mechanical device based on a second machine learning process and based on an aggregation of the machine data collected from the plurality of additional mechanical devices, wherein the plurality of additional mechanical devices and the mechanical device are each a same type of mechanical device;
• and
• providing a three-dimensional design environment associated with the three-dimensional model that renders the updated physics modeling data for the mechanical device.