soft logic in baseline

mimoune,



Nobody is perfect, not even the contractor scheduler, he migh have a valid plan that must be changed as things happens.



The argument that if the Contractor could not predict the future and that becuase he did not stuck line by line item on his original plan then his plan was wrong is in error.



You are correct in covering your ass, is part of your duties, by covering yourself you are protecting your client. My comment on the Contractor not being perfect and you knowing how to do it better was intended to highlight that the responsible for the planning is on the Contractor, that his plan in no way is perfect, that is dynamic, perhaps if we are lucky valid at one point in time because of changing conditions.



Because of this even contemporaneous methods are questioned. At times it happens "true" plan predicts a delay claim to be valid and then it happens through time schedule relationships changed and there was no such cause and effect link.



The reviewer must verify the plan as it being reasonable that is all he can do. Even when reasonable how can he discover that true plans by the Contractor is what he expresses in the submitted schedule?



What Gary suggested in his posting no 5 to me is the way to go when the Contractor uses soft links, although he suggests a note for allowing for these links be changed I believe this unnecessary, perhaps he is suggesting this as many of us have been victims of reviewers that insist that Contractor sticks to an inflexible plan. Yes soft links should be disclosed for a better understanding of the schedule. Even some people argue the resource leveling algorithm is a black box that hides cause and effect relationships, not sure but believe the forensic analyst can reconstruct these links in a credible way. Whether we like it or not we cannot escape to the reality these links can change, and most probably will, as the job progress. Allow the contractor to make necessary adjustments, your responsibility is for asking him for a justification whenever he changes logic.



At home, Owners require the Contractor to submit the original baseline and through the course of the job very rarely require for a revised baseline. Because of changing conditions the Owner avoids making reference to the Baseline schedule but to revised actual contract conditions. He avoids the implications a new baseline might represent, everyone knows if wrongly done it might implicitly change contract conditions in favor of the Contractor and against the Owner, it can be vice versa so everybody is happy and makes no mention to Baselines schedules. Schedule updates (not Baselines) are allowed to be modified but these modifications are required to be disclosed and justified. Schedule updates are never considered as implicit contract changes contrary to Contractual Baselines.



Best regards,

Rafael

We have two computation methods for PDM, we have continuous PDM and Non-Continuous PDM each have is rules about how the model works, is up to the person that creates the model to follow the rules.



While there are proponents that insist Non-continuous PDM is correct because it gives similar results to the traditional Activity-on-Arrow applications I believe them wrong, it hides the logic requirement of activities to be continuous, that if non continuous they are better modeled as separate activities and not in a single summary activity where the person preparing the model lose control on how to split and misses much of the information regarding the individual segments.



For this to happen laddering is not a requirement it can happen whenever you have a FF link, kind of the most prone link type to get the unwary scheduler in the trap when he wanted the model to perform in other way.







Activities do have float at the end and float at the start, what is missing is a metric that will provide us not only with a warning but how much activity duration can be increased before it starts to consume the available float to the right of the activity. Eventually someone will take the lead and provide us with the metrics; I challenge Asta to do it. I was suggesting for it to be named intermittent float, the difference between the ES of an activity as computed under Continuous PDM vs. Non-Continuous PDM in this way you decide on a one by one basis if you are to split the activity and in how many chunks, a decision not to be automated, especially not across the board as P3 does when you select to allow for activity splitting. I believe P3 still falls short of warning you when only FF predecessor link is present.



By the way there are different definitions for laddering, it can be a sequence of activities linked by SS and FF links where the end and finish links are not necessarily kept continuous with the activity duration. There is also a particular definition for laddering where the links are kept continuous by adjusting the remaining durations of the activities as progress is applied, the problem with this approach is that the amount of activities in any ladder is limited, no stairway to heaven allowed, and cannot recall if there were other limitations regarding the lag values. In the construction industry the latter can be a limiting factor while in other occasions it can be a requirement.



Best regards,

Rafael

Hi Andrew



I have yet to come across a baseline set up in a true ladder format.



Any combination of SS FF +>- lags then Yes all the time.



If I did have to review a programme that replicated your scenario - which I have tested - then my first choice would be to use the "delete & link around" icon.



Best regards



Mike Testro.

Gary,

The first sentence should read:



The ability to ignore a link type when rescheduling was put in v9, (I think), of the software to correct a particular error in the software when resheduling SS - FF ladders

Mike,



It’s one to watch out for.



Most often come across it when impacting delays into a programme using fragnets - where some of the activities are either initially set to zero duration, durations become zero during the analysis or activities inserted as milestones, (all of which could be for many valid reasons).

Hi Andrew



Thanks for showing us another good reason why ladder logic is not appropriate for construction programmes.



Best regards



Mike Testro

Gary,



The ability to ignore a link type when resourcing was put in v9, (I think), of the software to correct a particular error in the software when sheduling SS - FF ladders.



Try this:



Act 1 Dur 5 days

Act 2 Dur 5 days

Act 3 Dur 5 days



Link 1 to 2 and 2 to 3 with SS(0) and FF(0) and reschedule.



All should currently have the same start and finish dates



Now make Act 2 Dur = 0, (as if it never happened), and reschedule



Act 3 start date now incorrectly equals Act 1 finish date as the software reschedules on the Act 1 FF(0) logic link to Act 2.



Change this link to a different type of link and set the software to ignore that type of link - reschedule and the start date of Act 2 is now correctly the same as Act 1.



That’s what that little function is primarily for!!!!

Gary,



Thanks for your exposition, is very clear and agree with it.



About switching off and on of links in Spider you can do this in several ways you can toggle them at the links table by switching the “Use in scheduling” field, with the use of formulas you can create sets for easier switching. The advantage of this way is that the link will still be displayed although toggled off, is transparent, you don’t lose track of it.







You can also use several Reference Books, the Spider way to edit database tables. Here you can change the values of the “Use in scheduling” field or delete and restore the links.



Here I will not be surprised if my clients insist in using soft links; I will follow their wishes and will check against the resource leveling after switching off the soft links.



Anyway, both software are providing much needed functionality in this regard.



Best regards,

Rafael

There’s also a feature in Powerproject to ignore certain types of links (eg resource links) when you schedule.

Though I had a play with this feature the other day, and I’m not convinced it works properly.







Mimoune: I would advise rather than insisting soft links are removed from the baseline programme, request that all soft links are noted, either via a user field in the software, or on a seperate sheet submitted with the baseline. This gives you the ability to understand the ’true’ critical path e.g. for EOT analysis without placing unecessary constraints on how the contractor can run his project.



May also be worth giving the contractor a waiver to change these noted soft links as he sees fit throughout the life of the project, if such activity is not normally allowed by the contract.

Unless soft links are reviewed & ammeneded regularly to reflect changing resource requirements, they will quickly distort the schedule into something far from optimum.



Cheers,



G

Hi Rafael



In powerproject you can set up different coded links - for instance a resource driven link could be a blue dashed line while a logic driven link would be black solid.



(red and green are reserved software colours)



When the links are set to be curved when vertical the visual impact is immediate.



Best regards



Mike Testro

Dear Mimoune,



Level 4 is a Control Schedule that you request every 2-3 weeks.



You need to make sure that the Plan is correct and the schedule is correct. That means all the components making the schedule are constructed properly.



Asking your subcontractor to submit the resources is essential. Showing the soft logic sounds good. You need to review the complete schedule with everything associated with it.



With kind regards,



Samer

Mimoune,



If the contract is vague, then you have not many options other than to call it to the attention that although not specifically prohibited you reserve the right to remove the soft links in your evaluation of EOT’s for an unobstructed view on cause and effect.



Remember that most software do not provides true float after resource leveling and the only practical option the scheduler has to see true float is to create these links and resource load the schedule as to verify his assumptions as to the “soft links” without running the resource leveling calculation. It is not a bad idea under these conditions. In a way it is also good when the software algorithm does not follows prioritization. Another reason is that few or maybe none of the commercially available software provides you with all these links, these can become many and although can be considered redundant for the calculations of float they are not redundant for the analysis of your driving and near driving links.



I have mixed feelings about it and for the moment do not rule it out as a practical approach to tame the issue, at least not until I see a functionality that shows me all links. For example I might have two limited resources being used by two activities ending at the same time, these two resources will be required by a third activity at the same time, then both activities at the moment are resource driving the third, I should be able to see it. And what if one activity is to end 2 days ahead, the second can be driving the start of the third activity but the first do have a soft link with resource float, again I should be able to see it. Perhaps for the moment the only way to see the links is through the use of these soft links and the resource leveling algorithm just a help to assist you into their creation. A good resource leveling algorithm is still of great help as you can always keep two sets of the same schedule, one with the soft links and another without them, it would be easier if at a click of the mouse you could toggle on and off these soft links.



The Contractor keeps control of the means and methods; the Owner is entitled to be informed as to minimize any impact his actions can have on the job. If the Contractor Plan is reasonable it should not be rejected because you are perfect and know how to do it better.



Best regards,

Rafael