Forensic delay analysis with an automated simulation engine.

Hi Anders

Congratulations on the launch - it has been a long time coming.

I have developed a simple programme based on Lotus 123 which provides similar results on delay events.

I have been using it for the last 15 years but I keep it to myself.

I still have to do a formal delay analysis on the barcharts because tribunals will not accept the results at face value.

Good luck with the launch.

Best regards

Mike Testro

Thanks Mike.

This is written in Java. It is, I believe, not a black box so much as a transparent perspex box, well-grounded in solid mathematics and engineering simulation theory. For this reason I am confident that courts and tribunals will eventually accept its results. Is my confidence misplaced? Time will tell. We are all used to copious amounts of coloured bar charts, but I don't believe it always needs to be that way.

Getting all the data validation right was probably the most time-consuming and tricky part of the development. For instance, you have to ensure that there can never be any open ends, ever, at any instant of the simulation. You also have to ensure that no permutation of logic links can ever be entered that causes loops or cycles - otherwise the simulation will crash. 

We also looked at whether or not to have resource levelling and if so, what resource levelling process to use. The final result is a compromise: manual resource levelling. It tells the user where resource capacity levels are violated, and leaves it to the user to correct the violation manually in the next simulation.

Kind regards

Anders Axelson.