Using Markov Decision Process for Construction Site Management in Cameroon

Building construction sites are enclosed and restricted environments with minimal space to maneuver objects. Thus the organisational management of cost control, time and quality of construction projects is a huge challenge. Methods of project management commonly used by companies in developing countries are essentially planning schedules. Often, these methods are based on the assumption that complete knowledge of the project is known. Managing a construction project does not only depend on site factors but also on external factors such as different stakeholders involved on the project and weather conditions. As a result construction projects are often exposed to many vulnerable conditions leading to risks and uncertainties that generate decision problems affecting decision-making. In order to help project managers to optimize the construction management of the project, interactions or relationships between different vulnerabilities in a construction site have been identified and categorized. Based on the categorization, a methodology based on the Markov Decision Processes (MDP) was developed for optimising construction site decision-making affected by vulnerabilities. MDP are powerful analytical tools used for sequential decision-making under uncertainty, widely used in many industrial and manufacturing applications.