Q. The "muddiest" item in this module (a little contradictory to
my clearest item) was figuring out if MS Project will calculate the overall
critical path time for you. In fact, I still haven't found a way - can Project
do this automatically or do you need to add up the durations of the critical
activities yourself?
A. I use the dates, then compare for different options, as in crashing. For
the simple problems we do, it's easy to add the days on the critical path by
hand. You can also generate a report go to VIEW>REPORTS>OVERVIEW>Summary
and look at the durations.
Q. There is one part of networking that I do not fully understand the planning
method for. How do you add tasks that have no predecessors or successors to
a network diagram?
A. If it is AoN you have to add some sort of "Start" node. If it is
AoA, there is always a start node.
Here is a good narrative that points out that real life is more complex than
the simple examples given in the book.
There is one aspect of the CPM that hasn't been addressed by this chapter at
all. Unfortunately, in real life this constitutes the major drawback of the
CPM. When I was working as a planning engineer, updating and revising schedules
of large projects with durations exceeding three years, these schedules used
to have approx. 4000 activities.
The critical path might have run through 300 activities. Unfortunately, the
critical path kept changing on a week-to-week basis. The reason being that even
on projects of such long duration the difference of being on or off the critical
path is exactly one day, because, by definition, activities on the critical
path have to have a negative free float.
Long critical path chains keep changing constantly (e.g. a supplier informs
you that he can deliver earlier than expected, etc.).
Now, as a planning engineer you have the obligation to raise the flag whenever
something becomes critical. Unfortunately, nobody will really pay much attention
to the red flag if it keeps changing from week to week (e.g. this week race
boring is critical, next week it is the construction of the pressure shafts,
etc.).
So, again, the longer and larger the project, the more fluctuations in the critical
path.
Another problem is that the critical path does not show what is really relevant for the project and what not (e.g. a critical path of a hydro power project might run through the erection of a guard house or final landscaping; - of course this is obsolete). Might try a summary schedule that combines the major events.
So, what is the repercussion of all this? In our case (on a project), we spent
more time analyzing and re-directing the critical path than purely updating
the project. Part of the analysis was to foresee the future development of the
critical path and to analyze its stability and reliability.
The book does not spend a second addressing these considerations.
Further, a time scheduling often cannot reflect "partial realities". What I mean with this is that an activity linked with a FS-relationship needs to be 100% ready before its successor can start. Very often activities are only 80 to 95% finished before their successor starts. A SS-relationship with positive lag or a FS- relationship with negative lag would not do the job neither because the original intention of finishing the activity is valid until time pressure comes up and overlapping must take place. This is not the same as compression. Very Good.