Flexit: Flexible Location Encoding for Item Access in Medication Management Systems

Abstract

In hospitals and pharmacies, the demand for efficient medication management (e.g., item storage and access) is rapidly increasing, and the scale of medication management systems is gradually expanding. Intelligent medication management systems based on Internet of Things (IoT) can enhance the system efficiency and largely save labor costs through contactless medication dispensing and robotic item access. A typical medication management system that can store medicines usually includes several cabinets, each cabinet contains several drawers, and each drawer contains medicine boxes. In this paper, we propose, Flexit, a framework of flexible location encoding schemes that can encode the locations of cabinets, drawers, and boxes for efficient item access in an intelligent medication management system. Flexit includes a human/user interface, a central control unit, and a medication storage system, built upon medicine cabinets. Given storage space constraints, we propose a hybrid data compression method that combines trees and bitmaps, which is further optimized according to Huffman coding principles. Additionally, given a location code, the user/robot is capable to decode a number of locations, and Flexit employs a location-aware path planning algorithm to help access the stored items at these decoded locations at the minimal time-distance combined cost. We evaluate the performance of the data compression and path-planning algorithms through experiments over datasets, which show superior performance compared to existing methods, i.e., the compression algorithm can save 5-10 times the storage space on average, and the path-planning algorithm can reduce the system’s runtime by 20-50%.

Keywords

• Capacitated vehicle routing problem
• data compression
• genetic algorithm
• Huffman coding
• intelligent cabinet
• Internet of Things