A Critique of
Vahid Khodakarami, Norman E. Fenton, and Martin Neil
Article
“Project Planning: Improved approach incorporating uncertainty”
Table of Contents
Abstract
Introduction
CPM and Uncertainty
Bayesian Network
Conclusion
Abstract
Because of the unique nature of projects, there may be uncertainties or differences in the products, services, or results that the project creates. By its very definition a project and its activities can be new to members of a project team, generally necessitates more out-and-out planning than other routine work. It inevitably involves uncertainty. The basic inputs of time, cost, resources for planning are not deterministic and can be affected by uncertainty. Vahid Khodakarami, Norman E. Fenton, and Martin Neil feel there is a causal relationship between these uncertainty sources and project parameters and this causality isn’t considered in current project planning methodologies. In their paper they present an approach, using Bayesian network modelling that they feel addresses both uncertainty and causality in project management. And they show this using the Critical Path Method (CPM) to handle uncertainty and how you can manage different sources and parameters of uncertainty in project planning.
Introduction
It is well known that there are many techniques Project Managers can use to support better project planning. Critical Path Methodology (CPM) is one such tool that is used by many Project Managers. There are many ways to consider uncertainty in project planning. Group Creativity Techniques such as brainstorming, nominal group technique, idea/mind mapping, and affinity diagram, and multicriteria decision analysis all help to take uncertainty into consideration when planning a project. Khodakarami feels that these techniques rarely quantify uncertainty effectively. Their paper focuses on project scheduling and in risk management.
Risk management, while it includes identifying, analyzing and responding to potential risks by using mediation planning, it was felt that it pays too much attention to the positive events while minimizing the adverse events since the term risk is associated with events instead of uncertainty(Khodakarami, 2007). In their article the authors point out that using Bayesian Networks in conjunction with CPM helps to address uncertainty in scheduling because it allows you to take into consideration many different knowns rather than speculating on unknowns.
CPM and Uncertainty
Risk management is highly important to project planning. If the PM doesn’t consider and plan for risks they’re only planning to fail. The current view of risk is restrictive when it comes to uncertainty because it looks at events rather than sources of uncertainty. One source of uncertainty is activity duration. Normally duration is determined by looking at past history and by talking with subject matter experts (SME). Much uncertainty comes from not knowing exactly how long it could take to actually complete a task. This uncertainty arises from not knowing resource availability, possible occurrence of identified risks, lack of prior experience and the subject nature of the data such as expert judgement.
Critical Path Methodology is a deterministic technique for determining scheduling. It calculates the critical path (CP) by outlining the network of dependencies between tasks and determines duration for each arriving at what the earliest time is the project could be completed. It has a starting task and determines the order of the tasks using dependency between the tasks drawing out a network diagram that shows the tasks and their dependencies. Using the estimated durations for each task the PM can determine the earliest and latest start times and finish times for each task. From this calculation the PM would be able to draw out the critical path because he would be able to see which set of tasks have no slack time between them. This is the critical path of all tasks that have to be completed with their given timeframe, all other tasks have flexible amount of time between them that should they start earlier or later it will have no effect on the overall project schedule.
The critical path can be calculated using the following formulas:
D – Duration
ES – earliest start time
EF – earliest finish time
LF – latest finish time
LS – latest start time
The earliest start and finish times are determined by working forward through the network diagram, determining the earliest time a task can start and finish using the start and finish time of its predecessors. The first task always starts with zero. The latest start and finish times can be determined by working backwards through the diagram.
Bayesian Network
Bayesian Networks provide decision-support for a wide range of problems involving uncertainty and probabilistic reasoning. It is a probabilistic graphical model that represents a set of random variables and their conditional dependencies. For example, it could represent the probabilistic relationships between diseases and symptoms. If symptoms are known, the network can be used to determine the probabilities of other various diseases (Wikipedia, 2015).
An example of a Bayesian Network could be:
Sprinkler < – > Rain
Wet grass 
The main use of Bayesian Networks (BNs) would be for statistical probability outcomes. The idea is that the PM has a set of known events and they want to determine potential outcomes. BNs can explicitly quantify uncertainty, reason from effect to cause as well as vice-versa, make predictions from incomplete data, and combine objective and subjective data arriving at decisions based on visible auditable reasoning. resources’.
As the project advances duration of an activity becomes a known and by equating actual with predicted durations, the BNs can update the probable estimations for risks and resource usage and as such, this distribution can be used for later for estimating more accurately other activities with similar dependencies.
Conclusion:
Using BNs to help improve the planning process makes sense. It is better able to determine the probability of events occurring. Many current methods are too deterministic in their evaluations; CPM for example. But by combining CPM and BNs the PM can bring the power of BNs so that their plans schedules are more realistic and they’re better able to make an informed decision based on better information. But one has keep in mind from decision making under uncertainty is the risk that the project manager wishes to incur (Kerzner, 2009). Even with BNs being a better tool to use to narrow down uncertainty, it is still uncertainty.
References:
Bayesian network – Wikipedia, the free encyclopedia. (n.d.). Retrieved October 8, 2015, from https://en.wikipedia.org/wiki/Bayesian_network
Kerzner, H. (2009). Project management: A systems approach to planning, scheduling, and controlling. Hoboken, NJ: John Wiley & Sons.
Khodakarami, V., & Fenton, N. (2007). Title: “Project Scheduling: Improved approach to incorporate uncertainty using Bayesian Networks. Project Management Journal, 38(2), 39-49.
Project Management Institute. (2013). A guide to the project management body of knowledge (PMBOK guide), fifth edition. Newtown Square, PA: Author.
