An Overall Policy Decision-Support System for Educational Facilities Management: An Agent-Based Approach
Although K-12 public school facilities infrastructure investments are second only to highways, schools continue to suffer from an approximately $38 billion annual funding gap. Massive reductions in funding are forcing school districts to make tough decisions to optimize maintenance expenditures. Over the last three decades, a huge body of research has determined that the condition of school facilities do affect student health and performance, and some have further demonstrated that schools are overwhelmed by deteriorating facilities that threaten the health, safety, and learning opportunities of students. The currently available educational facility management approaches oversee the influence of the complex and mutual interactions between a school facility and its occupants. This thesis aimed to develop an overall decision support system for decision-makers that promotes efficient planning and management of educational infrastructure system by embracing a proactive management style rather than reactive. The proposed system consists of three main components: (1) an overall condition prediction model for educational facilities as a whole, (2) a tactical level Agent-based model (ABM) for classroom interaction simulation, and (3) a strategic level ABM for maintenance budget allocation. ABM was selected for its flexibility, natural representation of the problem, and suitability for modeling real-world complex systems with heterogenous agents. The first tool was accomplished through the development of a three-stage condition prediction methodology. The first stage aims to recognize the deterioration pattern of the educational facility as a whole by utilizing a Markov chain modeling approach. The second stage focuses on determining the overall useful service life of educational facilities. The third stage identifies the higher and lower limits of the educational facilities’ deterioration rate. The resulted model can help decision-makers plan and forecast their maintenance needs and better manage the available resources. The proposed methodology can be applied to any multi-component asset. The second tool, the tactical level decision support ABM, was developed to provide decision-makers with new insights into the effects of different maintenance polices on the educational system. The model simulates day-by-day classroom interactions and highlights the importance of preventive maintenance on the educational system’s major stakeholders (agents). The third decision support tool presented in this research is the strategic level model for testing the effects of different maintenance budget allocation strategies on the school district revenues, overall performance, enrollment size, and land values over years. ABM enhances the overall comprehension of the current situation and its complex relations, increases resource allocation efficiency, highlights the important factors affecting the system that are overlooked in traditional management styles, thereby improving the quality of educational outcomes. The main challenge in developing the proposed ABM was identifying and quantifying the main stakeholders’ complex interactions due to the uncertainties inherent in human behavior. This thesis demonstrated the need for a holistic bottom-top asset management modeling approach rather than asset-centric top-down approach. The case study results of this research confirmed that ABM has great potential as an asset management tool for decision-makers that can provide a comprehensive and holistic understanding of the system dynamics.
Kandil, Purdue University.
Education finance|Education Policy|Civil engineering
Off-Campus Purdue Users:
To access this dissertation, please log in to our