Under resource constraining the software simply figure it out the re-sequencing for you and if good software will disclose the new resource dependencies [the dashed links in case of using Spider Project] it will automatically re-align when needed. Keep in mind some software are incapable of disclosing resource dependencies as a different type of link.
If your software is incapable of showing resource dependencies as links different to hard link maybe it will be able to show you the delay reason as some software can do, otherwise you should consider looking for better software capable of disclosing what have been done.
Even in case of no changed conditions you want to know what activities are delayed because of resource constraining, software incapable of disclosing this is very poor and will force you to use archaic methods.
It seems to be the claim of client that he is finding out your solution is far from good, otherwise if your re-sequencing is good there is no need to object to manual methods unless specifically prohibited by the contract specifications that might prohibit use of soft logic instead of automatic resource leveling. I use automatic resource leveling but believe the ones in charge of managing the schedule shall be free of using their preferred method, still scheduling shall be managed, it is not anarchy.
The problem I find here is that you are a Subcontractor telling the General Contractor how to reschedule when it should be that he as a GC takes charge, manage the schedule and require you to follow it, then the impact, if any, you shall be able to claim. I do not see the GC taking the leadership role or maybe it is he is not managing a CPM at all. Maybe he did not took into consideration your resource limits before contracting your company and pretends you provide unlimited resources, now he is changing the contract conditions and shall take responsibility as a GC to manage the schedule and assume responsibility on the impact.
I would prefere also to use "manual" change in the logic instead of automatic leveling of resorces even if it is not simple.
Theexample I proposed, in reality, is a summary and each building is composed by, at least, 30 activities.
To complicate the picture there are also (in tha "actual" situation) partial access dates to certain areas of the building allowing to start somewhere with some activities but not enough to complete the all building.
It seems, in any case, any possible solution pass trought assignment of resources to the programme.
This would allow to compare if a different distribution give right to extra-money but extention of time is difficult to prove and calculate simply mainteining the logic and links of the baseline and simply change acess dates.
Please bear also in mind that we don't have yet an approved baseline programme and this discussion is in place in order to find an agreement on the programme itself.
Your and others commentor help has been really appreciated.
Sorry to come late into this discussion but you are facing a typical Sub-Contract claim situation.
The solution is application of an As Planned v As Built analysis.
Youre baseline is driven by the agreed planned access dates.
These have to be replaced by the As Built access dates and the programme rescheduled.
Now - if the delay can be mitigated by redeployment of key resources then this must be taken into account.
For instance if the original deployment was for block A > B > C > D but the access dates were A > D > C > B then you are obliged to re deploy your key resource to the new sequence - provided always that it did not cost you any money.
This is achieved by adjustment of the soft logic - Resource levelling does not enter into the argument.
Rafael + Vladimir and I have had this discussion over many years but this is the first time it has cropped up in an delay analysis topic.
The reason why not to do it is that once you have pressed the "resource level" button it is very difficult to find out what your software hase actually done - andif you don't know then how are you going to convince a Judge?
When you adjust the logic you can write down what you have changed and what the result is for every change.
Keep it simple.
Best regards
Mike Testro
Member for
21 years 8 months
Member for21 years8 months
Submitted by Rafael Davila on Sat, 2014-05-24 21:28
Revised schedule using preferential logic - for illustration purposes only access to Building A was delayed.
Revised using resource leveling and no preferential logic - because the software I use can display resource dependencies I am displaying them as dashed arrow, these are not hard links and whenever the resource leveling figures out there is a better sequencing they will be changed.
As you can see the job without preferential logic ends at month 11 instead of month 14 that results if keeping original preferential logic.
If the special lifts and resources are not continuously required, then there is a possibility that impact is less than under the assumption that they are needed across the board during the whole work on each building. If this is the case more detailed schedule and detailed resource loading is needed for better planning.
Member for
21 years 8 months
Member for21 years8 months
Submitted by Rafael Davila on Sat, 2014-05-24 12:35
When you said - "All building are similar to each other" I interpreted it as the Teams can be interchanged but if not then let is be as per actual requirements.
That Teams can not be perfectly doubled means they shall be modeled as different resources as you suggest but the artificial links shall be eliminated, the resource leveling will create the appropriate resource dependencies during resource leveling. Then when things change these resource dependencies might change and the resource leveling algorithm will do it right or at least feasible as there is not such thing as an assurance you will get optimal schedule even when using software capable of looking for optimal schedule, something P6 cannot do as they use a rudimentary algorithm that is very dependent on user input.
In the example you provided Buildings A, C and E require Team 1 and are linked sequentially but if you keep the links and delay Building A the other two will be delayed while in reality you can start with a different sequence that would require changing the links. Eliminate the artificial links and allow the resource leveling algorithm do its work, it is better than manually figuring it out on every update on jobs with hundreds/thousands of links.
Many schedulers know that leveling algorithms, even the poor ones, will usually provide better schedule than manual methods and this is why they require you to avoid using preferential links to drive resource leveling.
In this case a wise scheduler from the Owner's side would ask you to keep schedule as directed, using his own calculations without use of such artificial links and prove it can be done more efficiently without resource conflicts than with keeping initial preferential links. It is the subcontractor responsibility to follow the GC schedule and if his solution have lower or no impact you got to follow his schedule, the Master Schedule.
That the schedule was accepted does not means the GC cannot order changes to the schedule. That he did not noticed the use of preferential logic does not means he have to live with it, especially if the implementation of preferential links is poor and not adjusted as needed.
Beware that the combination of preferential logic and resource leveling can also be a trap.
if I understand well t would be better to load resources on programme and avoid the links.
In that case, with different access dates, we would have a different distribution of resources (maybe extra-cost) but, probably, no right to ask for an extension of time.
This is what we wanted to avoid for several reasons:
Teams can not be perfectly doubled. Some resources needed for special lift, to be done in each locations, are costly and not easy to find in the local market (i.e. strand jacks, special trailers, huge crane). In this respect we can not accept that principle as the programme (without link from building to building) could potentialy show several lifting with strand jacks ("special operation" in terms of resources and procurament time of equipment) at the same time and this is not possible to us.
Than, as you can imagine, 9 girls cannot do a baby in a month and we would not like to accept the principle that our contract duration can be "legally" squezeed putting all the activities in parallel at the end of our contract period.
At least the baseline should reflect a picture wich is the one in the subcontract agreement and approved considering the information we had at the time it was signed (including amount of resources that we planned to use considering a certain total duration).
Than, if access changes or are delayed, accelarations, time and cost impacts (if any) should be assested.
Is this position correct?
Many thanks again for your help.
Member for
21 years 8 months
Member for21 years8 months
Submitted by Rafael Davila on Fri, 2014-05-23 16:37
The use of preferential logic instead of true resource leveling is tricky.
Because the work on the buildings is independent of one another if you delay Building A it does not mean an automatic delay on any of the remaining buildings you artificially linked using preferential logic as to force the elimination of resource overloads. To avoid this trap many specifications prohibit the use of preferential logic as a substitute of resource leveling, therefore try with the following.
Keep baseline schedule as is.
Remove the artificial links between the different building activities as they do not logic depend on one another and keep only the link to finish milestone.
Update start no earlier than constraints equal to revised changed access dates.
Because work on buildings are similar then team 1 and team 2 are equally capable of working at any building, define them as same resource for which 2 are available. If your software provide true skill modeling and multi-resources then better use skills on multi-resources; P6 lacks skill and multi-resource functionality but in this case you can do without it.
In case two buildings are available and the two teams are also available but you defined for them the same team the resource leveling will delay the second while in reality both can be worked at the same time using the two available teams.
Execute a schedule run using resource leveling.
The delay caused by GC will be the difference between actual baseline and revised leveled schedule finish milestone. There is a possibility the delay will not be as much as if you keep the artificial or preferential links and no wonder the GC is complaining, your update adds to his own delay.
Member for
21 years 8 monthsUnder resource constraining
Under resource constraining the software simply figure it out the re-sequencing for you and if good software will disclose the new resource dependencies [the dashed links in case of using Spider Project] it will automatically re-align when needed. Keep in mind some software are incapable of disclosing resource dependencies as a different type of link.
If your software is incapable of showing resource dependencies as links different to hard link maybe it will be able to show you the delay reason as some software can do, otherwise you should consider looking for better software capable of disclosing what have been done.
Even in case of no changed conditions you want to know what activities are delayed because of resource constraining, software incapable of disclosing this is very poor and will force you to use archaic methods.
It seems to be the claim of client that he is finding out your solution is far from good, otherwise if your re-sequencing is good there is no need to object to manual methods unless specifically prohibited by the contract specifications that might prohibit use of soft logic instead of automatic resource leveling. I use automatic resource leveling but believe the ones in charge of managing the schedule shall be free of using their preferred method, still scheduling shall be managed, it is not anarchy.
The problem I find here is that you are a Subcontractor telling the General Contractor how to reschedule when it should be that he as a GC takes charge, manage the schedule and require you to follow it, then the impact, if any, you shall be able to claim. I do not see the GC taking the leadership role or maybe it is he is not managing a CPM at all. Maybe he did not took into consideration your resource limits before contracting your company and pretends you provide unlimited resources, now he is changing the contract conditions and shall take responsibility as a GC to manage the schedule and assume responsibility on the impact.
Member for
11 years 5 monthsThank you Mike for your
Thank you Mike for your help.
I would prefere also to use "manual" change in the logic instead of automatic leveling of resorces even if it is not simple.
Theexample I proposed, in reality, is a summary and each building is composed by, at least, 30 activities.
To complicate the picture there are also (in tha "actual" situation) partial access dates to certain areas of the building allowing to start somewhere with some activities but not enough to complete the all building.
It seems, in any case, any possible solution pass trought assignment of resources to the programme.
This would allow to compare if a different distribution give right to extra-money but extention of time is difficult to prove and calculate simply mainteining the logic and links of the baseline and simply change acess dates.
Please bear also in mind that we don't have yet an approved baseline programme and this discussion is in place in order to find an agreement on the programme itself.
Your and others commentor help has been really appreciated.
Member for
19 years 10 monthsHi GiordianoSorry to come
Hi Giordiano
Sorry to come late into this discussion but you are facing a typical Sub-Contract claim situation.
The solution is application of an As Planned v As Built analysis.
Youre baseline is driven by the agreed planned access dates.
These have to be replaced by the As Built access dates and the programme rescheduled.
Now - if the delay can be mitigated by redeployment of key resources then this must be taken into account.
For instance if the original deployment was for block A > B > C > D but the access dates were A > D > C > B then you are obliged to re deploy your key resource to the new sequence - provided always that it did not cost you any money.
This is achieved by adjustment of the soft logic - Resource levelling does not enter into the argument.
Rafael + Vladimir and I have had this discussion over many years but this is the first time it has cropped up in an delay analysis topic.
The reason why not to do it is that once you have pressed the "resource level" button it is very difficult to find out what your software hase actually done - andif you don't know then how are you going to convince a Judge?
When you adjust the logic you can write down what you have changed and what the result is for every change.
Keep it simple.
Best regards
Mike Testro
Member for
21 years 8 monthsBaseline Schedule.Revised
Baseline Schedule.
Revised schedule using preferential logic - for illustration purposes only access to Building A was delayed.
Revised using resource leveling and no preferential logic - because the software I use can display resource dependencies I am displaying them as dashed arrow, these are not hard links and whenever the resource leveling figures out there is a better sequencing they will be changed.
As you can see the job without preferential logic ends at month 11 instead of month 14 that results if keeping original preferential logic.
If the special lifts and resources are not continuously required, then there is a possibility that impact is less than under the assumption that they are needed across the board during the whole work on each building. If this is the case more detailed schedule and detailed resource loading is needed for better planning.
Member for
21 years 8 monthsWhen you said - "All building
Member for
11 years 5 monthsThank you Rafael for your
Thank you Rafael for your prompt reply.
Your suggestion seems clear:
if I understand well t would be better to load resources on programme and avoid the links.
In that case, with different access dates, we would have a different distribution of resources (maybe extra-cost) but, probably, no right to ask for an extension of time.
This is what we wanted to avoid for several reasons:
Teams can not be perfectly doubled. Some resources needed for special lift, to be done in each locations, are costly and not easy to find in the local market (i.e. strand jacks, special trailers, huge crane). In this respect we can not accept that principle as the programme (without link from building to building) could potentialy show several lifting with strand jacks ("special operation" in terms of resources and procurament time of equipment) at the same time and this is not possible to us.
Than, as you can imagine, 9 girls cannot do a baby in a month and we would not like to accept the principle that our contract duration can be "legally" squezeed putting all the activities in parallel at the end of our contract period.
At least the baseline should reflect a picture wich is the one in the subcontract agreement and approved considering the information we had at the time it was signed (including amount of resources that we planned to use considering a certain total duration).
Than, if access changes or are delayed, accelarations, time and cost impacts (if any) should be assested.
Is this position correct?
Many thanks again for your help.
Member for
21 years 8 monthsThe use of preferential logic
The use of preferential logic instead of true resource leveling is tricky.
Because the work on the buildings is independent of one another if you delay Building A it does not mean an automatic delay on any of the remaining buildings you artificially linked using preferential logic as to force the elimination of resource overloads. To avoid this trap many specifications prohibit the use of preferential logic as a substitute of resource leveling, therefore try with the following.