Own a system for all time your Cash Loan Company Cash Loan Company faxless hour online website.

Does Reliability Matter in Project Planning?

by Hal on April 21, 2009

in Last Planner, Theory of Constraints, project planning, project scheduling

If you're new here, you may want to subscribe to my RSS feed. Thanks for visiting!

Two standard six-sided pipped dice with rounde...
Image via Wikipedia

Ihad an interesting question about plan reliability. "Why does reliability (PPC) matter?" My first thought was, "Where do I start? Of course it matters!" Ok. Breathe in; breathe out. I know what to do.

A good planning system will enable project team members to fulfill their promises just as they make them

Let's start with PPC. We recommend measuring planning reliability using the measurement percent of promises completed. Our thinking is if people can do what they promise to do, then the planning is good. It doesn't mean that the future should be just as we planned it to be. Life's not like that. But, if we're doing a really good job with our planning, then most of the promises we make for completing work will be kept. PPC is a measure of reliability.

Reliability is an indirect measurement of the quality of the planning system. One of our goals of planning is to accomplish what we set out to do. Another way of saying this is, "I promise to get "x" done by time "y"." A good planning system will enable project team members to fulfill their promises just as they make them. Is this important? (I'm thinking to myself, "Hell yeah!") Reliable completion of our promises releases work for other performers. By keeping my promise, the next person in line is able to keep their promise. And so on, and so on. That is the nature of projects.

When task completion is unreliable, then the completion of the project is indeterminable

Eli Goldratt tells an instructive story in The Goal on the nature of the compounding of dependence with variation. It's a little game (using match sticks) that shows that when one action depends on a prior action that variance in the performance of those actions leads to an unpredictable completion of the chain of actions (covariance for the Six Sigma folk). In other words, when task completion is unreliable, then the completion of the project (chain of tasks) is indeterminable. That's why reliability matters. Being reliable on a task-by-task basis is the only way we can get our projects done on time.

Reblog this post [with Zemanta]

{ 13 comments… read them below or add one }

1 GLen B. Alleman April 21, 2009 at 9:15 pm

Actually it’s not reliability – it’s credibility. That’s what we’re after on on NASA and DoD programs. If the Integrated Master Schedule (IMS) and the associated Cost model is “credibile” we can then assess the impact of these baseline items on the deliverables.
The statement “I can get you done by time X” is meaningless in the absence of the statistical confidence interval on the date X.
Reliability is meaningless if the schedule is not credible – meaning it is not a schedule that results in a desired outcome. The assessment of schedule (and cost) credibility is the primary focus of the Program Planning and Controls activities on all DCMA (Defense Contractor Management Agency) audited programs (anything > $20M).
The Journal of Cost Analysis and Parameters speaks to this nearly every issue. There are scheduling based journals with the same focus.
Give me a “credible” schedule anytime over a reliable one.

2 steve April 21, 2009 at 11:50 pm

Yes ofcourse ! reliablity matters a lot in project planning ,because risks/reliablity decide the outcome of any project !

3 Hal April 22, 2009 at 1:38 pm

You don’t have to have one without the other. By “credible” I think you mean that it is “believable.” I don’t think you mean “accurate” or “correct” because schedules are about the future. We encourage project teams to continuously evaluate their schedules for believability. If anyone has doubt about the plan we ask them to speak up. Reliability gets at another matter. We can agree that we should do whatever we’ve come up in our plans. But that doesn’t mean that people are capable in the moment of getting something specific done. Yet, when one person is late in performing, that can cascade through the whole project. It is therefore critical that we have a process that assures that what should be done can be done. We call this a “make-ready” process. It’s also know as “no excuses” planning. By addressing the constraints to the “should” plan we create the conditions so it “can” be accomplished. This process produces the basis for reliability.

It turns out our plans are no longer credible when unreliability in task completion occurs. Can’t have one without the other.

4 Mike Sullivan April 24, 2009 at 12:30 pm

You said something that I felt the need to expand on further. ”When task completion is unreliable, then the completion of the project (chain of tasks) is indeterminable.”  Folks tend to get stuck thinking about the project through the critical path.  This behavior is easy to predict as it is taught in scheduling class, through software and reinforced through conversations and planning.  It is common to hear superintendents say that they are focusing on the important activities first during the week. Once those are “under control” the focus is shifted to those not so important tasks. This process leads to but one outcome, non-critical tasks are not focused on until they become critical. Activities stack up on the job until everything is now critical. The project is now overloaded with workers. As the demand on resources increases, delivery time increases as well.

The focus is on the critical path not on individual tasks. Folks will argue that the critical path is the (chain of tasks) and therefore the focus is in the right area. A chain with links linked end to end is the longest path from point A to B. Our focus should be to help others see that if the links are connected end to end and top to bottom, they turn in to chain mail which is the shortest path from point A to point B.

5 Glen B. Alleman April 26, 2009 at 3:09 pm

This is one of those “semantics” discussion.
By “credible” in the DoD and NASA world, it is meant that the plan, schedule, cost, and other “attributes” could actually be completed in the manner stated in the baseline contract. Doesn’t mean they will. But that they “could.” “Could” is the first step to “can.”
What this paradigm does, is wrap all the point description – reliability, repeatability, any of the “illities” into a synergistic whole and present it as “credible.”
It is credible they could could construct a 4 person spacecraft and fly bck and forth to the Station by 2011. “It is credible that we could take off the shelf hardware and software and upgrade all the voice, data, and video systems in the USAF for under $9B.” Stuff like that.
In the absence of the “credibility” test, the next level down – the “illities” have no home in which to be tested.
Accurate is a meaure but ONLY with the variance. No point estimate can be useful in the absence of the statistical variance. This approach is mandated through DID 81650 for all DoD and NASA programs and is being applied to all FAR procurements in 2009.
I see you term “reliable” to mean “acheivable.” This woudl be wrapped in the “credible” description. NASA actuall uses the “credbile” term for assessing the proposal to include all attaributes of the offer and once awarded a continuous assessment of the credibility on a monthly basis.
What’s interesting between construction and others, is this notion of having the right people in the right place, at the right time. I read quite a bit about this issue in construction. The resource management activities on a defense program focus heavily on marshling the staff for the job. Doesn’t mean it always works, but there are dedicicated people doing that job every week.
So “It turns out our plans are no longer credible when unreliability in task completion occurs. Can’t have one without the other.” Start with a credible plan, then execute that plan within the variance limits and the reliability issue is addressed.

6 Hal April 26, 2009 at 4:00 pm

Point of clarification:

I refer to reliable in terms of doing what we say we will do. We are only reliable when we make a promise and then fulfill it. Reliable is an assessment. Like all assessments it is made based on past actions with a concern for the future. I might have a history of reliably completing my promises, however what we know about the future could have us make a different assessment. For instance, I might deliver at the very last moment. Future tasks might involve more complexity than those I historically completed. You might assess me as not reliable for the tasks that are coming.

7 Glen B. Alleman April 26, 2009 at 4:17 pm

Using your first sentence, the assessment of reliable happens only after you’ve done what you said you were going to do. Reliable is a post hoc assessment. Post Hoc “refers to looking at the data—after the experiment has concluded—for patterns that were not specified a priori.”
We only know if your’re reliable after the fact – as you say. Credible is an assessment that can be made a priori. The assessment is performed prior to the execution.
I’m suggesting – from external guidance at NASA, NAVAIR and Air Force – the “credibility” of the plan is a superior (higher) attribute in which reliability of the past performance is one of the indicators of credibility. The assessment of the credibility of the plan, cost, and technical solution can be peformed analytically in the absence of execution and performed during execution using the “past performance” data from the project AND the continued assessment of the future plans for cost, schedule, and techncial performance.
As you state reliability is important. My OP suggested that credibility would be a better place to start, then flow down to relaibility and many other attributes. In the absence of credibility, the assessment of reliability, is a trailing indicator – “so far so good.”

8 Gregory Howell April 26, 2009 at 6:42 pm

What does “credible” mean in this context? Credible to whom? For what purpose? One of the first steps in a project is determining in a very rough way if there are sufficient time and resources. Sometimes this takes more detailed planning and sometimes not – particularly if there is no great pressure on either or similar work has been completed. Even so different people will have different views – there is no firm standard. This is particularly true in exploration where the primary asset may be resourcefulness rather than resources. Was the discovery of the new world by Columbus credible? Finding the path to the Pacific by Lewis & Clark? That Lean Construction would change the industry?

I am confused – is “credible” a paradigm? Glenn is missing the point here. Reliability is always within limits understood even if not expressed. We are talking here about reliable promises in projects, specifically design and construction work. These aren’t estimates for some higher authority. Rather they are the promises one crew leader makes to the next. They are based on the judgment of the person making them, their assessment of the situation. The likely variance could be expressed, minus or plus half a day – but they usually aren’t and that would be reasonable.

Does the DOD or NASA ever review those initial credible schedules to determine the source of incredible overruns. Maybe a look at task reliability would help.

9 Dennis Stevens April 26, 2009 at 9:23 pm

This is not just a semantic discussion. This discussion highlights different philosophical points of view in the planning and managing projects. While reliability of promising can impact the credibility of a plan, there is a clear distinction between credibility of the plan and the reliability of the promising.

A credible plan is one that is capable of being believed. This is an attribute of the project plan itself. Given a reasonable set of conditions this project should be able to be performed in this way according to this schedule for this cost. On the other hand, reliability of promising is an attribute of behavior on the project. Specifically regarding how often the interactions between performers happen as promised.

Both are important. We need a credible plan. A credible plan gets the project funded in the first place. A lot of focus in project management goes into developing and governing the credibility of plans.

But, reliable promising gets the project delivered. In Glenn’s world, this reliability may be perceived to be at the contractual level – the focus is on the plan. Hal is talking about how the reliability of individual interactions impact an organizations ability to fulfill its contract which the impacts the ability to deliver the project – the focus is on behavior.

This distinction is important. I can’t get better at delivering a project by improving the credibility of the plan. I can drive dramatic improvement in the performance of the project by improving the reliability of promising within the project organization.

10 Glen B. Alleman May 1, 2009 at 6:23 pm

Credible to only person that matters – the customer

11 Glen B. Alleman May 1, 2009 at 6:37 pm

I understand your paradigm is based on making promises and keeping them. I’l not challenging that. That’s important. All I’m suggesting is that behaviour is subordinate to those promises being based on a credible plan so when they come true the project provides the value expected by the customer.
In our project world where “emergent” is the basis of all we do, reliability is less important than have a credible plan on how to get to the end.

The promises represent a statistical attribute – measured after that fact – of 1,000’s of activities on a program in our domain. Those promises exist inside the larger framework of an end to end credibility assessment that is performed offically monthly in the NASA 533M report and weekly on Thursday for the Control Account Managers meeting. So Greg, yes NASA and DoD review these activities weekly.

My only suggestion from Hal’s OP was that reliability was necessary but not sufficient for project success. You can have reliable delivery of individual work efforts and the project itself be late and over budget, because each of individual reliable activities does not represent a credible plan for success. Formally in the DoD/NASA/DOE world the Plan is the Strategy for successfully delivery the project. The schedule is the seqeunce of activities needed to execute that Plan.
Once we have a “credible” Plan, the execution of the Schedule of that Plan, needs to be “reliable.” So in fact you can improve the probability of success P(s), by improving the credibility of the Plan. The inverse is true well. I can decrease the P(s) for the project by decreasing the relaibility of the individual work processes inside a credible plan.

This is the paradigm of the Integrated Master Plan / Integrated Master Schedule (IMP/IMS) mandated for all Federal programs > $20M.

12 Proworkflow January 2, 2010 at 12:50 pm

It is common to hear superintendents say that they are focusing on the important activities first during the week. I am confused – is “credible” a paradigm? Glenn is missing the point here. Reliability is always within limits understood even if not expressed. We are talking here about reliable promises in projects, specifically design and construction work.

13 planning projects April 9, 2010 at 8:32 pm

Of course, reliability matter in project planning. How come it wouldn’t be? It is important and crucial to the result of the project.

Leave a Comment

You can use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>

Note: This post is over 5 years old. You may want to check later in this blog to see if there is new information relevant to your comment.

Spam Protection by WP-SpamFree

Previous post: Have Project? Get Yammer

Next post: Beginner’s Mind Leads to Mastery