Optimization and Rule-Based Models for Hospital Inventory Management

Abstract

This thesis shows how optimization, rule-based models, and operational analytics can be used to help manage hospital surgical inventory. The models were created for AITA™, a team under Johnson & Johnson’s Ethicon subsidiary. The AITA™ Smart System is an intelligent inventory management solution that stores, organizes, and distributes products via Kiosk, Smart Shelf, and Mobile Hub devices. Every device requires a planogram, or a visual representation of which products to stock and the location of each product. This project focuses on creating models to automatically build and update these planograms. The models presented in this paper have already been adopted by the AITA™ team and have begun to show accuracy and efficiency gains when compared to the current manual process. Model-designed kiosks cover, on average, 7% more historical procedures than hand-made kiosks. Also, model-generated planograms are free from manual product selection and sorting errors. From an efficiency perspective, automatically creating and updating planograms will save the AITA™ team an average of 145 hours annually for every hospital served. These accuracy and efficiency gains will add value across the entire chain of care. The AITA™ team will have more time to grow their business and to develop new features. Meanwhile, providers will save time when managing and retrieving hospital inventory, which will free up more capacity for direct patient care.