Finding the Nearest Wheelchair: the Application of AI Search Methods to Logistical Problems in the Hospital

Abstract

Ever since the invention of the Global Positioning System (GPS) and its slow entry into the public domain in the 80s and 90s, location tracking has become an increasingly important tool for scientific research, surveillance, and commerce. The potential for increased efficiency and quality of care has resulted in an increase in Indoor Positioning System (IPS) technology uptake by the healthcare sector. The main focus of this master s thesis has been on the application of search methods from Artificial Intelligence (AI) literature to a logistical problem in the hospital, facilitatedby the use of location tracking technologies.Using a Requirements Engineering (RE) framework the thesis explores use cases where time spent looking for wheelchairs can be reduced. The three use cases: View Wheelchair Location , Find Available Wheelchair , and Find Nearest Wheelchair was identified, where finding the nearest wheelchair was found to be most suitable for further investigation. The thesis proposes a search framework for describing the problem of finding the nearest wheelchair given the current position and task information of a hospital porter. Furthermore, the RE study proposes a set of requirements that the search framework must adhere to. A literature study of AI search methods such as the well known Dijkstra s and A* algorithm along with different kinds of acceleration methods were performed, and a prototype of the search framework was built. The search framework and all the proposed search methods have been evaluated by performing a series of experiments on two different kinds of datasets made by the author.