Greetings
From all above discussions what is the conclusion. What is prefered to use SS, FF with lags + or - or using FS with - lag.
Please this important because the consultant ask me to change all FS - lag links to SS and FF.

CARLOS,

DPH ACTIVITY TYPE
There are four types of project activities:
•Duration;
•Productivity;
•Hammock; and
•Milestone.

An activity of Duration type: duration is the most basic activity information when the activity is included into the schedule. Duration does not depend on resources assigned to the activity.

An activity of Productivity type: initial data for this activity type is its volume and assignments productivity. Activity duration is calculated during scheduling.
LAG

In addition to the type of logical relationship (link) between activities (Finish – Start, Finish – Finish, Start – Finish or Start – Start) the user can specify a lag:
•Time lag – time period between the moment when requirements (preceding event) are accomplished and successive event (e.g. successive activity start or finish) can occur;
•Volume lag – specifies the amount of work that should be done on the preceding activity to allow the successive activity to start.

When updating activities with time lag if the time has been consumed then the start of the successor activity can happen just at DD irrespective on how much work have been performed. Therefore if I start an activity but progress is slow the software will tell I can start successor when it is not necessarily true. With the use of volume lag the successor activity will be delayed to the projected time remaining volume lag is projected to finish.

I assume the software will determine the projected time to end of volume lag as:

= (remaining volume lag)/(team volume production rate)

This must be confirmed by Vladimir as I have not been in need to use the functionality. This I believe is very powerful when you use Fragments with different volumes, resource assignment levels and productivities as to get appropriate scaling as these will determine activity duration and projected (but not fixed) volume lag duration. Fortunately the computer does the number crunching.

A few formulas speak a thousand words. Hope Vladimir will clarify Duration of Volume Lag Volume Formula and Activity Duration Formula for DHP type "Productivity" Activities. At times I get lost but with a few sentences Vladimir makes everything crystal clear.

If you are interested in an in depth discussion of volume lag, when ready you can issue a new thread for this purposes at Spider Forum. You bet I will be most interested into the issue. There we can discuss sample jobs for this purpose, it will be my learning experience as well.

Carlos I am thinking on a sample job of two activities linked by volume lag, the first crew working 100% of the time on the first activity while the successor activity crew will be producing twice the rate but working 50% of the time. A volume lag equivalent of one day production of the feeding activity. Just let me know.

Best Regards,
Rafael

Mike,

When not installing, the rebar crew is preassembling footing rebar, column rebar and all they can preassemble. As soon as there is a distance for the crews to work without it be a safety hazard or a physical obstruction between the crews then the rebar crew will install the preassembled re-bar and here is where you got the volume lag into action, is not even a full day, but it happens. And excavation crew does not stop, next day they will continue with excavation while a crew gets the water out of the 2 hours excavations on prior day.

On small jobs excavation crew will assist the concrete pouring crew and will use the tire mounted backhoe to move the concrete vibrators, to move the toweling machines and so on.

We are in no need to create so many activities as long as SS and FF volume lag keeps true relationship for the start and finish of the activities. We can model what happens across several days or weeks with volume lags, no need to create hundreds of unnecessary activities to display the continuous (or at times intermittent) volume lag relationship. Is simple, activity durations can span several days or weeks while volume lag can be a small quantity that will represent perhaps an hour of work between the crews.

“The dry season is usually from November to May” is a lie, last week in the second half of the “Dry Season” we had several consecutive days with over 8in/day (20cm/day) rainfall in several towns. No need to call National Guard or FEMA, no need to ask for these areas be declared disaster zone.

Best regards,
Rafael

Mike,

Here our specs require us to keep under separate activities work performed by different crews, by different supervisors, by different contractors. I can understand you mix such activities as you are allowed to do so, you are free to consider such practice as “good practice”, but here it is not.

Is very common for the excavation be done by a separate crew than that who installs the reinforcing steel and the crew that pours the concrete. These activities are also considered separate as individually they have different predecessors, such as materials delivery among others, they start and finish with a lag relationship, not at the same time, and we must model the lag relationship that exists whenever they do and cannot pretend it does not. Even if the activities take a few weeks and the volume lag relationship is small compared to the volume of work, the relationship is still there and the progress of the successor is still continuously dependant on any differential progress of the predecessor.

As a matter of fact the need for feeding relationship is not necessarily limited to be a One Way street; it can be a two Way Street I believe it was modeled by Micro Planner many years ago with their implementation of ladder activities; it was not just SS and FF relationships with one way feeding.

Take for example placement of reinforcing steel as a continuous feeding operation that as it progress it opens for the electrical contractor to install embedded conduits and as the installation of conduits progress then it allows for the top reinforcement be installed. Under this functionality as far as I can remember a single one of the ladder activity can be the controlling driver for the progress of the remaining activities.

Best regards,
Rafael

Carlos,

The following is a small sample of many activities whose progress depends on the other on a continuous way.

Finish Grading Slab on Ground= = > SOG reinforcing steel
Footing Escavation= = > Reiforcing Steel== = > Concrete Pouring
Foundation Walls= = > Termite Treatment= = > Backfill immediately after treatment
Masonry Installation= = > Metal Frames to be installed and mortar filled simultaneously

If you do not overlap the activities efficiency will be lost.

Because it rains a bit here we must plan site work operations to be continuous, we cannot stop the continuity of the operations. Take for example footings, we excavate footings in the morning, place reinforcing steel cages previously pre-assembled and pour whatever was prepared during the day otherwise on the following day footings might be flooded and re steel with mud. There is no way we can predict the exact footings to be placed on a single day as to model the sequence in tandem per day, even if we could predict it would be impractical to set such a detailed schedule. It does not mean we will use a single activity for each operation across the job, we split the schedule in manageable segments in accordance to the network scale, but the continuous feeding relationship must still be modeled.

If you want a real representation of activity relationships then use of volume lags instead of time lags. Is very easy and better than breaking the sequence into hundreds of discrete chunks most probably will be different in the actual execution.

Best regards,
Rafael

I vote FS, long live to it. I’ve noticed field managers tend to abuse FF and SS, and use SF for "Planning backwards". I HATE SF links, I cannot even imagine an event requiring to start an activity to finish another. Maybe in industries different than construction.

In my (little) experience, I’ve found that Gary’s approach is not just highly effective, but natural if you are doing traditional CPM. There’s no other way to do it. Even when you start updating your schedule, you’ll end breaking the activities into pieces. That is, if you will be updating the schedule.

There are 2 factors you must face to understand why FF and SS produce bad schedules:

1. You must understand the use of Float. It is not just a plain number, it is the result of the CPM. Being the result of an algorithm, it is made to simplify your life. Float is like money, it is not good or bad, all depends on how you use it.
2. You must UPDATE your schedule. You’ll find yourself breaking large activities into smaller, more manageable chunks.

The point is, remember that the plan is to be executed, not just to be on the paper. Your duty is to make plans and execute them, not to make schedules.

Best Regards,
Carlos.

Rafael,

Accepted volume lags would often solve this problem (though in my experience, many SS+ relationships aren’t there becuase of lack of a volume lag functionality, but becuase of lazy scheduling)

If you don’t have volume lags, it is simple to use just FS relationships to do the same thing.

e.g. B starts after A is 20% complete:
1)Split A into 2 activities: "A 1st 20%" and "A remainder", linked by FS
2)Link B by FS to the 20% activity.

If you use SS+ relationships a lot, they will require a lot more memory power / better programme notes to ensure they are not overlooked during updates. It’s just not a robust method.
And of course some clients won’t let you change lag values once they’ve approved the baseline!

Cheers,

G

Abed,

Best practise is to use neither.

The major problem with SS+ is that once the predecessor has started, how it progresses has no bearing on the schedule

The 2 major problems with FF- are:
1) You don’t know when the sucessor should actually start until the predecessor has finished, by which time it’s already too late.
2) A change in the duration of the sucessor will have no impact on the end date.


If you insist on using one of these, I would suggest SS+ is the least worst option.

Here we go again

Only use FS lags.

Best regards

Mike Testro