Dear Professionals,
How to develop a revised baseline schedule using last update? I need to do the following:
1. Need to, prepare the revised baseline schedule, by taking as-built data of last updated schedule.
2. Need, Planned Value (PV) of revised schedule to be exactly equal to Earned Value (EV) of last updated schedule.
3. Need to, Remove all actuals from revised baseline.
Your response will be highly appreciated.
Regards,
Ali
Dear Experts,
I would like to thank;
Rafael Davila
Vladimir Liberzon
Johannes Vandenberg &
Zoltan palffy
For all your valuable suggestions and support regarding my query.
I have summarized your suggestions below, which i have utilized to prepare my schedule. I want you guys to just review for any other future refinements.
SUMMARIZED SUGGESTIONS A & B are mentioned below:
A. PREPARE REVISED SCHEDULE FROM LAST UPDATEDSCHEDULE:
1. Copy the last updated schedule to a save place.
2. Correct all Out of Sequence logic relations
3. Correct all Un-satisfied relations.
4. Note in schedule narrative for all logic changes.
5. Adjust the remaining duration of in progress activities, the relations of not started activities and include any new scope if applicable.
6. Use following global changes:
a. If activity Status = Completed, Then Original Duration = Actual Duration
b. If activity Status = Completed, Then Planned Start & Planned Finish = Actual Start & Actual Finish date respectively.
c. If activity Status = in progress, Then Original Duration = Actual Duration + Remaining Duration
d. If activity Status = in progress, Planned Start = Actual Start & Planned Finish = Finish
7. Get this Revised schedule to be approved by project team & competent authorities.
8. After approval, Maintain and assign baseline (save a copy baseline option)
B. DE-STATUS REVISED BASELINE SCHEDULE
1. Make a copy of the scheduling model and saved it under the same EPS with different ID.
2. Add the following two columns in activity layout.
a. Primary Constraint
b. Primary constraint date
3. Define Mandatory start constraint for all started activities. This can be done by sorting activities by activity status (in-progress & completed) and defining the mandatory constraint at the top of the schedule and use the fill down facility for all the other activities.
4. Make two new columns in the layout i.e.
a. Actual labor units
b. Actual non-labor units
5. First remove all the actuals from the actual resources. This can be done by deleting the resource at the top of the schedule and use the fill down facility for all the other activities.
6. Choose Tools, Global Change and run the following Global Changes. zero.
a. If: Resource Type equals Material, Then: Actual Material Units = 0
b. If: Activity Status is not equal to Not Started,Then: Activity Status equals Not Started
c. If: Activity Status is equal to Not Started,Then: Remaining Duration = Original Duration
7. Set the data date equal to the Project Start Date and schedule. To find the Project Start date, choose Enterprise, Projects. Click on the Dates tab and note the Planned Start date.
8. Calculated the schedule by pressing F9 and verify the final results.
Ali,
What you have done and what Vladimir propose I find a very good alternative, the client should waive the requirement for use of date constraints on started activities. I works for purposes of EV, not necessarily for purpose of an As-built schedule. Yes it saves a lot of time, As-Built logic is not needed to get good results for EV purposes.
About Global change for Original duration = Actual duration I believe that for it be right for started but not finished activities it should be something like OD = Actual + remaining, I do not recall the details on how P6 handles OD so you must check it yourself.
A sensible client should allow for this exception and not require client unnecessary burden for the purpose at hand.
Because of your right attitude it is very hard the project will end up in court and most probably never there will be the need for As-built logic, why waste time with it.
Your procedure requires changing the DD to project start while I believe Vladimir's does not, keeping visible the DD of Baseline Update, in addition Vladimir's procedure does not require fixing all OOS logic or fixing OD. Spider Project provides a links table where it identifies the links that are creating the OOS as Broken Links, here you can filter them all and at a single click delete those links.
EVM specification for purpose of Baseline revisions only should allow on the Revised Baseline non-contractual constraints and the keeping of actuals, opening the door for such pragmatic procedure.
Best Regards,
Rafael
Dear Vladimir,
Thank you for your quick / fast solution for OOS logic. It is really helpul and speedy to correct huge number of OOS activities.
I found total 48 OOS activities in my schedule log, which I corrected by using:
1. Changing most of OOS links from FS to SS relation including necessary lags.
Dear Rafael Davila,
Thanks for sharing such informative link regarding calculation of as-built critical path. Its worth reading it.
rafael you are right, client requires revised baseline schedule with As-built logic for all completed and in progress activities upto data date.
Regarding de-statusing the revised baseline, i have used following.
1. Mandatory start constraint option
2. Global change for, Orignal duration = Actual duration
3. Remove all actuals
4. Set data date to project start
Now, i have de-statused revised as-built baseline schedule, but when i remove mandatory constraints and calculate schedule, activities re-organize themselves as per defined logic.
I am finding it very burdensome to change all As-Planned Logic to As-Built logic. Thats why i have used mandatory constaraint, although it is not a good option.
Regards,
Ali
Rafael, I suggested to set SNET dates after actual information was removed and planned data restored. It shall be done not only for OOS activities but to all delayed activities. Yes, other way is to set lags but it is much more time consuming. Setting SNET=Start is applying one formula. That is why I called it easy solution.
Vladimir,
What you propose shall do with regard to EV but still is missing the requirement to erase all progress data and perhaps the erasing of all non contractual constraints if the contract prohibit use of date constraints other than the contractual.
It looks like the client wants you to create the revised baseline using "As-Built Logic" something that have been criticized for numerous reasons.
It is possible the DOD, the father of the creature [EVM], considers using As-Built Logic within their options to update the baseline. To me the only option that makes sense is erase all variance, the effort on the other options for a flawed methodology is not worth it. Erase all variance by using the updated schedule as the revised baseline is simple, is the only option one that do not require both parties to be in agreement, in DOD contracts as per their guidelines can be opted by contractor at any time.
CONVERSION OF AS PLANNED LOGIC TO AS BUILT LOGIC
As per AACE:
See also:
Seems like Ali solved OOS using lag.
Fortunately at home we do not use EVM for our jobs unless they are Federal Government jobs that require it, we do not have such issues with the Baseline Schedule.
Best Regards,
Rafael
Ali, an easy solution is to set Start No Earlier Than constraints to already started activities making SNET equal to Actual Start and removing incoming links for OOS activities.
Dear Rafael Davilia,
Yes exactly, OOS logic correction, was the main problem.
It feels really good when we trace and resolve a particular issue by going deep down:)
As far this schedule is concerned, this one is not using Resource Leveling function. It has Activty Type "Task dependant" and Duration Type " Fixed duration & Units" .
Thank You.
Best regards,
Ali
Ali,
You are getting very close, never give up. It looks like a 100% consensus OOS shall be investigated.
Two thumbs up for Johannes, this is the spirit of how we should debate the issues.
I would like to know if this particular schedule use the resource leveling function.
Best Regards
Rafael
Dear Johannes,
Thanks,
I will correct all OOS logics now.
You know, the only problem is with in progress activites.
I will correct them and check it out.
Really appreciate your continous support and guidence.
Ali
Hi Ali
In general, out of sequence "Work completed for an activity before it is scheduled to occur. In a conventional relationship, an activity that starts before its predecessor completes shows out-of-sequence progress". causes instable results in the calculations.
I suggest the all OOS activities are resolved before you reschedule or rebaseline.
Regards Johannes
Hello Johannes,
Thank You!
Yes, i have assigned the new baseline as you said, i am pretty sure i have done these steps correctly. I have followed exactly the same procedure as you suggested.
Difference between PV (of revised baseline) and EV (of last update) is equal to 3.22% (of budgeted total cost)
May be it is because, i have not corrected OOS (out of sequence logic) ???
Ali
Hi Ali
Have you assigned the correct baseline? You should have now two baseline. The old and the new baseline. If you still have assigned the old baseline then the values are different.
Regards Johannes
Could it be that the following reference helps?
I no longer use P6 but recall having my issues with period performance data, maybe if you run a "post period performance" or however it is called by P6 it could help to fix the EV differences. To get true EV curve you should do it at every update/version in sequential order. The same goes with regard to an ever changing baseline, on every update window everything changes.
With the use of a few global changes you might erase all progress data, including resource data. With the use of a few global changes you might adjust the revised durations of as built data.
So far so good, but as soon as you resource level the schedule to get a resource feasible baseline everything might fall apart. My experience has been that when things get complicated magically most gurus conveniently forget about resource leveling. Such is the case with Longest Path Theory that breaks down under resource leveling as acknowledged by Oracle.
As a reminder of some comments by Vladimir make sure you fix logic if some actual progress if out of sequence.
No matter what you do or what cumbersome procedure you invent EVM will still be flawed and keeping track of a misleading variance a nightmare.
If the client is happy with flawed half truth then make him happy, if he asks for nonsense give him nonsense, but always remember the flaws and limitations of EVM.
Good Luck.
First of all I thank you all,
Rafael Davila,
Vladimir Liberzon,
Zoltan Palffy,
Johannes Vandenberg
You guys are really helping and supportive. I really appreciate all your time and efforts.
1. You know, As per my client’s requirement,
PV (of revised schedule) must be equal to EV (of the last updated schedule) on the date of schedule revision.
For Example:
A project is delayed by two months having PV = $2000 & EV=$1500 on Data Date 5 Sep 2015.
Now, Schedule is revised on the same Data Date i.e. 5 Sep 2015.
So, at this specific date PV (of revised schedule) should be equal to EV (of last updated schedule) so as to correctly evaluate future planned values.
But still, even after using all suggested methods, planned value PV of my revised schedules is not matching with earned value EV of last updated schedule.
2. Also when I De-status the project, and calculate the schedule (F9) all dates changes.
As per my client’s requirement,
I need to submit De-status Revised Baseline (having no actuals)
Zoltan, I know that this is Oracle forum, so don't understand why do you repeat in each of your posts that this is not Spider forum.
If the project was executed then some activities were unexpectedly delayed, some were done out of sequence, some were delayed because required resources were not available, resource availability had been changed, etc.
Please explain how your solution takes all this into consideration. Are you sure that after final rescheduling that you proposed project dates and costs will be right?
What is a Half-Step Schedule? - “Bifurcation (a.k.a. half-stepping or two-stepping) is a procedure to segregate progress reporting from various non-progress revisions inherent in the updating process. Elements that are considered to be non-progress revisions include:
o Addition or deletion of activities
o Split or combined activities, using new activity IDs
o Addition or deletion of logic links
o Changes to lag value of logic links
o Addition, deletion or changes to constraints
o Changes to Original Durations (OD)
o Increase in Remaining Durations (RD) such that RD becomes greater than OD
o Changes to RD not accompanied by changes to Percent Complete
o Increase in RD of activities that have not started
o Changes to calendar assignments
o Changes to holiday assignments within a pre-existing calendar” [1]
Please note that Ron Winter procedure misses to address resource constraining and changes in resource quantity availability or staffing. Similar to AACE International recommended practice that always miss to consider resource availability.
A Base Line is essentially a set of dates and costs frozen at the start of the project and used as a basis for performance evaluation as the project progresses. Every time there is a new baseline everithing can change. I do not believe trying to make the baseline an "All inclusive" schedule is a good idea, as if CPM Baseline Schedules were resorts.
In any case even if you do not care about resource feasible plans it is too much trouble for a method that is flawed with regard to the matching of EV and critical path activities.
Vlad
This is a Oracle Primavera forum NOT a SPIDER forum and my solution will work 100% of the time not on a RARE CASES
and my solution is a practicale solution and is the only solution
Hi Ali
1. Need to, prepare the revised baseline schedule, by taking as-built data of last updated schedule.
I suggest the following procedure
2A Suggested next step
Revise the schedule to suit your requirements. What-Is, crashing or fast track
When the schedule is ready and approved the updated schedule shall again be baselined.
So, now you have three baselined.
3. Need to, Remove all actuals from revised baseline.
First remove all the actuals from the actual resources. This can be done by deleting the resource at the top of the schedule and use the fill down facility for all the other activities.
The next step is to remove all the actual finish dates. This can be done at the same way with the fill down option
The following step is to do the same for the actual start dates. This can also be done with the fill down option but now you have to confirm the removal every date with an mouse click.
Then set the new start and the data date and remove all the start en finish constraint.
The last step is to modify the new schedule to suit your requirements.
Good luck
Dear Zoltan,
As you suggested, I have copied the current project (with admin settings at completion values with current dates) but, faced a problem that, PV (of revised baseline) NOT EQUAL TO EV (of last updated schedule.)
Below is snapshot of a sample project:
Last updated schedule:
[[wysiwyg_imageupload:2585:]]
After assigning new baseline (copy of current last updated schedule):
[[wysiwyg_imageupload:2586:]]
Note that, PV is not equal to EV of last updated schedule
Can you please suggest, how to fix this issue? How PV of Revised Schedule will be equal to EV of last updated schedule?
Ali,
Your high level of education and manners is obvious, you are good.
To update baselines using P6 [instead of using Spider Project as an Add-in], use of add-in and auxiliary software is an option frequently used by many to supplement P6 lack of functionality.
You might consider the creation of "Half Step Baseline" as per following article.
Use of the P6 Maintain Baseline Utility allows the Scheduler to save copies of project schedules
as “Baselines”. Once progress has been applied to the schedule and an appropriate Baseline
Schedule from the past is available, one can update an existing Baseline Schedule. Proper use
of the Update Baseline Utility to create a Half-Step schedule, allows the Scheduler the ability to
effectively and efficiently evaluate the impact of progress versus non-progress revisions
between schedule updates.
I do not use Earned Value [EV] nor Earned Schedule [ES] as both are flawed, though I recognize sometimes you are required to show it as per contract requirements. I find all alternatives on how to update EV Baseline of not much value, including the "Half Step Baseline" as it does not solve the fundamental flaws in EV/ES. Fortunately the DOD does not recommends it's use in Fixed Price Contracts, no matter the $_amount.
Hope this help
Good Luck
Zoltan,
I suggested potential solution without going into details understanding that Ali does not use Spider.
You suggested another solution that may work in some special and rare cases.
Let's hope that somebody will propose real and practical solution.
Cheers,
Vladimir
Vladimir
As you have suggested to others if your answer is about Spider select appropriate forum
http://www.planningplanet.com/forums/spider-project
How to develop a revised baseline schedule using last update? I need to do the following:
1. Need to, prepare the revised baseline schedule, by taking as-built data of last updated schedule.
answer: copy the current schedule
2. Need, Planned Value (PV) of revised schedule to be exactly equal to Earned Value (EV) of last updated schedule.
answer: copy the current schedule
3. Need to, Remove all actuals from revised baseline.
answer: see link on how to destatus a schedule I have detailed nstruction there as a pdf
http://pmsite.com/forums/viewthread/3064/
Dear Vladimir Liberzon,
Rellay appreciate your response but, my company don't have Spider Project and i have never used it.
Is there any way to do it in P6?
Ali
Dear Ali,
I can explain how to do all this if to use Spider Project.
Can it help? One of the ways is to export the schedule to Spider, do it there and then export back to P6.
I expect that it also can be done by exporting data to Excel, changing data there and then importing back.
Best Regards,
Vladimir