Keywords: Critical path method, Monte Carlo method, probability, risk analysis, scheduling, simulation, stochastic models. 1. Ph.D. Candidate, Construction Engineering and Management Program, Department of Civil and Architectural Engineering, Illinois Institute of Technology, 3201S. Dearborn St., Chicago, IL 60616-3793, USA. Email: firstname.lastname@example.org
INTRODUCTION For generations, the design and construction communities all around the world have used extensively project scheduling systems such as Critical Path Method (CPM) and Program Evaluation and Review Technique (PERT). Commercial scheduling software have become very popular in the construction industry. While these software programs are generally well designed and do a good job supporting construction scheduling functions, they are mostly CPM-based and do not calculate probabilities of completion. While CPM has been used extensively due to its simplicity, PERT has seldom been used and only in very special cases. A network is comprised of activities that make use of resources and require time to complete. The total project duration is obtained by computing earliest start and finish times for each activity. The method of calculation is available in many publications (e.g., Barrie and Paulson; 1992). As a basic principle, it is obvious that the more resources are allocated to an activity, the shorter the time needed to complete it. While each activity requires diverse resources, time is the most important factor in scheduling. Efficient scheduling allows the contractor to meet the construction owner’s time requirement. It also ensures that the contractor has sufficient capacity to meet contractual requirements at the time of bidding, as shown in Figure 1 (element 3 of ISO 9001; Hoyle, D; 1998). While the resources allocated and the time needed to complete an activity behave reciprocally, what is questionable is the total time that a project takes to complete the critical activities and the critical path, and the amount of resources needed to be allocated to each activity. These concerns are adequately addressed by most CPM-based software packages on the market. On the other hand, none of these packages addresses the stochastic and random nature of a schedule.
While CPM is based on deterministic fixed terms,...