Guild of Project Controls: Compendium | Roles | Assessment | Certifications | Membership

Tips on using this forum..

(1) Explain your problem, don't simply post "This isn't working". What were you doing when you faced the problem? What have you tried to resolve - did you look for a solution using "Search" ? Has it happened just once or several times?

(2) It's also good to get feedback when a solution is found, return to the original post to explain how it was resolved so that more people can also use the results.

S- curves

16 replies [Last post]
Thu, 2006-05-25 11:34
Iftikhar Awan P.Eng.
User offline. Last seen 6 years 4 weeks ago. Offline
Joined: 20 May 2006
Posts: 40
Groups: Petro-Chem / Oil & Gas
Hi.

When we say S-curves, are we essentially referring to Earned Value S-curves of BCWP,BCWS & ACWP?

What are other popular S-curves that a planning engineer should know about?
Top

Replies

Fri, 2006-06-09 15:38
#1
Svein Myklebust
User offline. Last seen 15 years 31 weeks ago. Offline
Joined: 27 May 2006
Posts: 25
Groups: Petro-Chem / Oil & Gas
I don’t zero out anything - it was just my way of telling what I think about Primaveras allways-on-schedule-frontline.
The history is a importent part of the project.

I agree with the rest of your comments,
but this is in the "perfect" project.
My experience is that in real life,
things are not that organized.
Customers who knows best, despite what the reports tell.
Projectleaders that consider planners secretarys that know how to draw a Gantt.
Late input - if any.
Changes of dates, budgets etc.
To many projects and not enough time
etc. etc.

Sam
Top
Fri, 2006-06-09 14:58
#2
Aneesuddin Zubair...
User offline. Last seen 7 years 4 weeks ago. Offline
Joined: 17 Nov 2003
Posts: 42
Groups: None
Hi Svein,

I do not know why & how you zero out all the history to run the update. The fact of the matter is an update gives you the history of the work completed upto data date as well as the forecast of the remaining works.
No doubt about it, Planner should report the true status of the project, whether it is two weeks behind schedule or two days, but should not raise false alarm without analyzing the schedule on all the aspects. I repeat again that measuring percentage progress alone is not sufficient, a prudent planner would also analyze the criticalpath (Comparing the actual/current dates against the planned dates on critical path) to have a realistic view of the project.

Cheers,

Anees
Top
Wed, 2006-06-07 11:04
#3
Svein Myklebust
User offline. Last seen 15 years 31 weeks ago. Offline
Joined: 27 May 2006
Posts: 25
Groups: Petro-Chem / Oil & Gas

Something tells me have been usin P3 for a while Anees.
The philosophy seems to have rubed of :-)

To me its frustrating to zero out all history every time we run a update an let all be on schedule. I prefer to
learn from the past. If we are two weeks behind then I want the management and everybody else to see that we are two weeks behind, not to search for les float or cartoonlike deliverytimes.
But this is just differnt approaches to a common chalange.


For me it all comes down to the small pieces of the schedule when we are talking progress.
No task, on the lowest level in the system, should so big that the people that realy do the work can’t determine how much of the job is done and thereby be able to give both the physical progres and the expexted finish date, for this little bit.
Then the rest is math, decided from the manhour-budget.

The exeption is of course follow up activeties such as planning various forms of administration. These are allways on schedule since we dont’n know what to expext round the next bend.


Svein


Top
Fri, 2006-06-02 08:31
#4
Aneesuddin Zubair...
User offline. Last seen 7 years 4 weeks ago. Offline
Joined: 17 Nov 2003
Posts: 42
Groups: None
Measuring the percentage progress is only one dimension of the project monitoring and control. I agree with James that the absolute value is of relatively minor interest, whether the SPI is 0.84 or 0.81. Measuring the progress percentage would help in determining the project’s health, but that alone is not sufficient, the performance of project on the time scale (ie. Comparing the actual/current dates against the planned dates on critical path) should also be evaluated to have a realistic view.

Choosing the method of updating the project plays a very important role, updating the progress based on schedule percentage complete or remaining duration. The realistic update of the remaining duration will tell when the project will finish irrespective of how early the activities are started.

Afterall these are the tools that enables Planner to recognize problems before they become unsolvable. In essence, it monitors and controls the actual work to be done and the time and cost required to do it.

Anees
Top
Fri, 2006-06-02 03:50
#5
James Griffiths
User offline. Last seen 15 years 20 weeks ago. Offline
Joined: 19 May 2006
Posts: 435
Groups: None
Paul,

I agree whole-heartedly with your sentiments.

The aspect which most Planners may find frustrating is the myriad methodologies of monitoring project progress/status. Moreover, how many people have had the opportunity to see a long-term project all the way through, in order to determine how accurately a particular monitoring method performed - never mind having to actually choose one beforehand.

My own experience/forensic investigations suggests that, even the conventional Earned Value system can manifest markedly differing results, depending upon the precise methodology of calculating progress et.al. As such, one learns to accept the limitations and, as long as the summary-level conclusion is approximately correct, then the absolute value is of relatively minor interest (ie. could you really determine that your SPI is 0.85 as opposed to 0.81? Could you really determine whether your progress is 40% or 44%?). Moreover, using another conventional method, such as Milestones Achievement, can also result in differences, depending upon the criteria that are used. However, you are quite correct in that there is no problem in using two methods, in parallel, to help verify the project’s general health. Once again, however, the absolute value is of minor interest. It is the trend that is the most important.

Ultimately, the potential complexity of determining a project’s health becomes so overwhelming to most PMs that they go back to their "gut-feel", based on practical experience.......and this is something that we, as planners, need to gently draw them away from.

As an aside, I’m always interested to learn of different curves. I’ve not yet come across the "Committed Curve". Your enlightenment would be most appreciated.

Cheers.

James.
Top
Thu, 2006-06-01 19:48
#6
Paul Harris
User offline. Last seen 1 year 9 weeks ago. Offline
Joined: 11 Apr 2001
Posts: 618
Groups: Content Moderators, GPCCAR Editors
James

You make some valid points and I agree with them all, but have you missed the point that I did say that I felt the Committed Curve was most important curve and I did not say the only one?

It is best to keep an eye on many attributes when monitoring project progress. One should always use a cluster of tools to give a clear picture of what is going one. Different projects may need different tools and it is not a one stop shop here, one should tailor the tools to focus on the appropriate elements of each project. Using one tool only, such as just the Critical Path, or just the Earned Value or just the Committed Curve or any just and other single tool will eventually land the project controls in trouble. Also using the tool inappropriately will also not help.

When talking about curves, my second favorite tool is the envelope between the early and late levelled curve. If a project goes outside of this envelope (on the late end that is) then you probably have some real problems. It is unfortunate that may scheduling software packages do not create the Late Levelled curve by default and one has to mess around to get this curve out of them. Setting the Performance Measurement Baseline with these two curves in mind is another useful tool. A Committed Curved based on Early Start Dates only is destined to fail as one does not know how far behind the curve one can go and still be able to finish on time. It is important to ensure that task on the critical path are also started on time.

As I am sure it obvious to every one, that ensuring work has started is the first step in executing an activity, one cannot monitor progress unless work has started and any tool that provides a measure on what has started must be of use on the appropriate type of project.

The issue of starting an activity that one cannot finish on time is often related to poor scheduling and/or one of a number of other issues, that I am sure we are all familiar with. But starting later than planned will not help getting it finished on time, that is assuming one has a valid schedule. Maybe a little realistic scheduling and resource planning would be of assistance here, but the issue of delays to in-progress activities is a real problem in many projects, but they cannot be delayed if they have not started. Also one might consider Critical Chain Scheduling if the Intermittent Progress problem is showing up a lot on your projects. I found this technique particularly useful on one project I was involved on.

Paul E Harris
Eastwood Harris Pty Ltd, Melbourne, Australia
Planning and Scheduling Training Manual & Book Publishers & Consulting
www.eh.com.au
Top
Thu, 2006-06-01 08:33
#7
Daison Garvasis P...
User offline. Last seen 12 years 48 weeks ago. Offline
Joined: 24 May 2006
Posts: 15
Groups: Petro-Chem / Oil & Gas
That is excellent explanation made by James!!! Just starting something to cover up other delays and then leaving the newly started activity as "started" forever is the day to day construction trouble unless monitored properly. Curves would give you a trend on "where you stand" but the Critical Path Analisis must be the one giving you some insight on "when you will finish". And yes, you can drive the car only after it is finished.
Top
Thu, 2006-06-01 11:43
#8
Raja Izat Raja Ib...
User offline. Last seen 13 years 4 days ago. Offline
Joined: 2 Jun 2005
Posts: 299
Groups: GPC Malaysia, Petro-Chem / Oil & Gas
Hi, I agreed with almost every post thread "how to cover up the delay" but i still try to figure out how?(what i mean which area have to look ) Its looks very easy to catch up the progress by doing parallel non critical activities at the same time but do U think it maybe become worse if critical path interfacing by non critical activities, then we have to rework.

regards
Top
Thu, 2006-06-01 14:00
#9
James Griffiths
User offline. Last seen 15 years 20 weeks ago. Offline
Joined: 19 May 2006
Posts: 435
Groups: None
Raja,

The essence of the discussion has been around the use of S-curves to monitor the project’s progress, plus my minor digression based on Paul’s statement regarding activities starting on-time.

Simply: If an S-curve is populated with the full value at the "start" of an activity, then the PM may be tempted to purposely start an activity, earlier than programmed, thus claiming its full value of "progress", but without any intention of continuing to work on that task. However, tasks that have been previously in-progress may have been delayed but will not necessarily be visible on the S-curve. This is because you are monitoring progress on the basis of the "start" of an activity, and not on its finish or pro-rata progress. So, when the news is bad, you divert the client’s attention by showing him what a good-boy you are, having started all these extra activities (as shown on the S-curve). Unfortunately, the-truth-is-out-there. You might be able to hide it for a little while, and buy yourself some time - but it will always catch-up with you at some point. This is why, when reporting progress, you (if you’re the client) should always ask for an ETC. If the sub-contractor fails to provide an ETC, you could extrapolate one by using the historical data from the S-curve.....but if that is mis-representing the picture in the first place, then you’ll get cocked data.

In summary: There are numerous S-curve types that can be used to monitor progress....but they can probably all be fudged, fiddled or manipulated in some way without necessarily reverting to direct lies. Ultimately, the whole thing relies on honesty!

As regards running critical and non-critical tasks in parallel; by definition there should be no problem in performing both tasks together. If there is a clash, then each task should have been programmed on a finish-start relationship, thereby placing both of them on the critical-path in the first place.

Cheers.

James.
Top
Wed, 2006-05-31 10:30
#10
Aneesuddin Zubair...
User offline. Last seen 7 years 4 weeks ago. Offline
Joined: 17 Nov 2003
Posts: 42
Groups: None
I agree with James, the practicing of starting few activites to cover-up delays somewhere else is common in construction industry.
Top
Wed, 2006-05-31 09:57
#11
James Griffiths
User offline. Last seen 15 years 20 weeks ago. Offline
Joined: 19 May 2006
Posts: 435
Groups: None
Paul,

You made some interesting comments about "..activities..started on time.."

Starting an activity, albeit "on time" is of little value unless you can also finish it on time - using any form of S-curve to monitor overall progress (each type will have its pros & cons). The important thing, is that the value of the S-curve will give a reasonably good indicator of the trend of progress, howsoever the progress is being monitored. Thereafter, it’s quite straight-forward to extrapolate an ETC curve using historical data. After all, you can only use the building once it has finished; you can only drive the car once it has been built etc etc.

Many PMs love to quote "XXX activity has started", expecting to have the client jump for joy, but far too often the "It’s started" statement is being used to cover-up a delay somewhere else. An analogy is where some VIP cuts the ribbon on the "Opening" of a new road...but it takes another 3 months before anyone can actually use it!

Cheers.

James :-)
Top
Wed, 2006-05-31 05:42
#12
Paul Harris
User offline. Last seen 1 year 9 weeks ago. Offline
Joined: 11 Apr 2001
Posts: 618
Groups: Content Moderators, GPCCAR Editors
From my point of view the most important S-Curve is a Committed Cost curve where the full value of the activity is placed at the start of the activity.

Too many people concentrate on when activities finish. Knowing that activities have started on time is more important than knowing that is finished too late.

At least one can ensure that it is stared as soon as possible. Unless one ensures that the activity starts on time there is little hope of finishing on time.

The Committed Curve give you early warning of when work is not starting on time.

Paul E Harris

BSc Hons, Certified Cost Engineer, PRINCE2 Practitioner
Director Eastwood Harris Pty Ltd

Paul E Harris
Eastwood Harris Pty Ltd, Melbourne, Australia
Planning and Scheduling Training Manual & Book Publishers & Consulting
www.eh.com.au
Top
Fri, 2006-05-26 10:23
#13
Iftikhar Awan P.Eng.
User offline. Last seen 6 years 4 weeks ago. Offline
Joined: 20 May 2006
Posts: 40
Groups: Petro-Chem / Oil & Gas
Thanks Brian.

While I have not yet used risk analysis modelling, this inofrmation would be helpful in future. I am currently doing my masters in construction management where we were introduced to concepts of decision under risk and risk analysis. I hope to implement it in near future.
Top
Fri, 2006-05-26 10:13
#14
Brian Stephenson
User offline. Last seen 18 years 6 days ago. Offline
Joined: 26 May 2006
Posts: 4
Groups: None
If you run any form of risk analysis modelling on your plans i.e. 3 point estimating... factoring in uncertaincy and risk (as all good planners do), you can plot any of your key milestones on a confidence/time S curve.

We use Artemis and Predict toolsets for this.

Very useful if you need to convince your PM that it just can’t be done!!
Top
Thu, 2006-05-25 16:02
#15
Iftikhar Awan P.Eng.
User offline. Last seen 6 years 4 weeks ago. Offline
Joined: 20 May 2006
Posts: 40
Groups: Petro-Chem / Oil & Gas
Thanks for the reply Thanat.

What are other important S-curves that we as planners should know about?
Top
Thu, 2006-05-25 11:47
#16
Thanat Patta
User offline. Last seen 18 years 4 hours ago. Offline
Joined: 15 May 2006
Posts: 5
Groups: None
It’s not necessary. The S-Curve can be only BCWS, BCWP, or ACWP but the Earned Value refers to the analysis among those 3 curves.
Top