At most Gamma Irradiation facilities, a common part of the process that can be improved is product scheduling. This process is largely manual and relies on institutional memory and intuition. Often, there is a large amount of historical data that goes unused or can only be coarsely used to inform new schedules. Nordion has been working to simply the data analysis process as well as automate, through Machine Learning, a predicted schedule based on available products.

This process can be improved further by adding information to the system through Mathematical Modeling to better inform the Machine Learning training data to further construct a more efficient product schedule.

This talk will discuss the current scheduling practices, future capabilities, and how Mathematical Modelling would improve scheduling efficiency. I will outline the current state of our capability and the process to create the Mathematical Model, show examples of CAD products within the product, and the current state of product scheduling within a Mathematical Model.