Stochastic project planning
Report 2008.TEL.7287, Transport Engineering and Logistics.
Computer software is increasingly used to assist in the process of project
planning. As project planning gets more complex as the included tasks increase,
software can greatly aid in reducing complexity. Every project can be divided
in a certain amount of activities. As every activity has a variable execution
time, total project completion time is often uncertain. Furthermore, some
activities probably will be indicative for total project completion times.
These activities will form the critical path of the project.
In this report, the development of a generic simulation program is described.
The program can be used to run any kind of project several times in order to
identify likely completion times and critical paths. Execution times of every
single activity are varied with every run made using a newly defined
distribution function. The simulation is based upon a conceptual model, made
using the process interaction method. This method makes sure the conceptual
model can be easily implemented in an object-oriented programming language such
as Delphi, using the module TOMAS. The simulation program includes a real-time
animation, increasing insight in the project execution process and the
differences in critical paths during different simulation runs. Furthermore, it
provides an easy user interface to inspect activity properties during
After verification of the simulation, an experimental case is elaborated. The
case includes an aircraft maintenance program, including an extensive network
of activities. Using this experimental case, the value of using the simulation
program for project planning purposes is illustrated.
Reports on Transport Engineering and Logistics (in Dutch)
, TU Delft