In previous paper we introduced the Bayesian version of traditional infamous CPM. The approach combines the simplicity of CPM method with the power of Bayesian analysis. The result is a powerful quantitative approach for handling uncertainty in project planning. It can capture, quantify, analyse and handle different aspects of uncertainty/risk in project parameters (in this case we are concentrating on ‘time’) in a way that no other technique can do.

The central component of our approach is a Bayesian subnet associated with each activity/phase of project. The ‘Duration’ subnet for each activity is the fundamental part of the model. This Presentation describes the new version of duration model and some of its features.

The model has several advantages that accompany with Object Oriented concept of BN can provide a very powerful quantitative technique for project planning. The main features of this network are:

* Objective estimation

* Trade-off between ‘time’ and ‘resources’

* Learning from data

* Proper uncertainty management

Return to Conference Programme