W13_Anton_Comparison PERT and Monte Carlo
1. Problem Recognition, Definition, and Evaluation
On blog W12.3, PERT method to define schedule contingency refer to risk identified, percentile desired is 90%.
This week, using the same data, I would like to apply Monte Carlo simulation as comparison to PERT.
2. Development of the Feasible Alternative
Fajar (2008) stated: “ Monte Carlo simulation is a method for modeling and analyzing system which involving risk and uncertainty”. (p. 227).
And Sullivan (2012) said: “ At the heart of Monte Carlo simulation is the generation of random numbers” (p. 503).
Using Ms. Excel to generate random number between zero to one by use function NORMSINV(RAND() to performing 1.000 simulation.
3. Development of the Outcomes and Cash flows for Each Alternative
Schedule P90 from Monte Carlo with total duration = 260.51 day
4. Selection of Criterion (or Criteria)
From W12.3 blog posting Plot duration P 90 to schedule from PERT result of duration are 285.64 day. And from Monte Carlo are 260.51 day.
5. Analysis and Comparison of the Alternative
Both PERT and Monte Carlo method has different 25.13 days in result.
6. Selection of the Preferred Alternative
Follow Humphreys (2012): “PERT technique is largely historical. Monte Carlo is understood best and is most frequently used” (p. 346).
7. Performance Monitoring and Post Evaluation of Result
Also refer to Humphreys (2012): “ The biggest challenge remains, which is collecting the best data to use for any of these techniques” (p.346)
8. References :
- Fajar, Adnan (2008), Aplikasi Simulasi Monte Carlo dalam Estimasi Biaya Proyek, Jurnal SMARTek vol.6, No. 4.
- Sullivan, William G., Wick, Elin M., Koelling, C. Patric. (2012), Engineering Economy. Chapter 6 Pp. 211 -228, Fifteenth edition, Prentice hall.
- Humphrey and Associates (2012), Project management using earned value , Chapter 27. Earned Value, Second Editions, Humpreys & Associates, Inc.