Using a predictive model to determine virtual usages of storage spaces in a storage infrastructure

A system for rendering in a computer display a virtual usage of storage spaces in a storage infrastructure, comprising:

• a processor;
• and
• a computer readable storage medium having computer readable program code embodied therein that executes to perform operations, the operations comprising:
• providing information on physical boundaries and usages of the storage spaces in the storage infrastructure, and a predictive model used to determine storage spaces for items to store in the storage spaces, wherein the predictive model implements machine learning;
• receiving, by the predictive model, inputs comprising physical boundaries of a selected storage space, attributes of users of the selected storage space, and attributes of items stored in the selected storage space;
• processing, by the machine learning of the predictive model, the inputs to output a virtual usage of the selected storage space within the physical boundaries of the selected storage space, wherein the virtual usage is defined by coordinates in the physical boundaries of the selected storage space and a virtual item to store in the selected storage space;
• generating, in the computer display, an augmented reality representation overlaying the virtual usage at the coordinates within a view of the selected storage space;
• and
• training the machine learning of the predictive model to output virtual usages in a positive training set for inputs in the positive training set comprising physical boundaries of storage spaces, attributes of users of the storage spaces, and attributes of items stored in the storage spaces.

A method for rendering in a computer display a virtual usage of storage spaces in a storage infrastructure, comprising:

• providing information on physical boundaries and usages of the storage spaces in the storage infrastructure, and a predictive model, implementing machine learning, used to determine storage spaces for items to store in the storage spaces;
• receiving, by the predictive model, as inputs, physical boundaries and usages of a selected storage space;
• processing, by the predictive model, the inputs to output a virtual usage of the selected storage space within the physical boundaries of the selected storage space, wherein the virtual usage is defined by coordinates in the physical boundaries of the selected storage space;
• generating, in the computer display, an augmented reality representation overlaying the virtual usage at the coordinates within a view of the selected storage space;
• maintaining historical virtual usage outputs including inputs to the predictive model, output virtual usages of the storage spaces based on the inputs, and rankings of appropriateness of the output virtual usages for the storage spaces;
• and
• training the predictive model to output the output virtual usages for the storage spaces in the historical virtual usage outputs with confidence levels based on the rankings of appropriateness for the output virtual usages in the historical virtual usage outputs.