deliver on time with the critical chain aka "theory of constraints"

The method of the critical chain or "theory of constraints" is coming from manufacturing and it deserves very close attention. I attended a presentation of Eric Robien on the topic at PMI France Sud and how it applies to project management and found it quite interesting. I herewith brush you a minimalist portrait of what I retained.

The critical chain, based on the theory of constraints, proposes new rules of management for the access to the really critical resources of the project to improve their global performance.

As reminded to us by the presenter, the current operating mode is often one of:

• To respect the overall timeline, the management insists that every task respects the schedule
• To respect these commitments, every person in charge of tasks adds a safety margin on each of his/her tasks
• The progress is measured according to the expenditures (time and investments) as compared to what remains to be done
• The schedules are more and more detailed ending in a myriad of individual tasks
   

But does this really improve the global results of the project?

Not always, and an inverse effect is even very often obtained.

First of all because we cannot ignore that a task project is difficult to estimate. It is often required to produce a very unique deliverable which was never realized previously.

There is thus a huge difference between a realistic estimation and reality of execution.

We have to recognize that the approach supported by the current operating mode drives us to build very important safety margins as we accumulate margins taken at the level of each and every task.

And moreover, these margins are wasted!

Indeed, Éric reminded us some of the classic syndromes which generate the waste:

• The law of Parkinson consists in using all the time granted to a task.
• The syndrome of the calendar consists in waiting for the scheduled date to start a task even critical.
• The syndrome of the student: conscious that he has more time than necessary in front of him to realize a task, the student will put back the start of it to make the other things which interest him much more.
• Multitask which provokes loss of concentration due to frequent stops and starts.
   

What is this new operating mode? What does the Theory of  Constraints propose?

Let us start by focusing on the identification of the critical resources (the weak links, the bottlenecks, the waiting lines which get longer and longer) and let's calculate for every project, deliverable or task which burn these resources, the value they return.

In product marketing, the best product would be the one who generates most revenue per hour of critical resource. Similarily, in project management, the best schedule is the one that will optimize the use of these critical resources.

Then, let's refocus on the important thing in scheduling which is to respect the overall timeline/date.

To do this, we shall:

• integrate into the critical path the resources constraints.
• reduce the safety margins of every task and
• mutualize the safety margins in a project buffer and auxiliary plugs.
• also, revise the sequencing of the tasks and projects to minimize multitasking
• and we need to implement a new follow-up methof for the project which will reflect our new approach.

The main indicators are:

1. The percentage of progress on the critical chain
2. The rate of consumption of the buffer of the project
3. The ratio between the above two indicators.

Naturally, this rather radical modification of our approach will not be achieved without real management of the change. The most often met resistances will be the management culture, i.e. a management based on distrust and non acceptance of uncertainties. The management has to focus on the execution of the tasks as quickly as possible and no longer to "only" respect the schedules.

Some internet links for a deep dive:

http://www.tocc.com/TOC%20for%20Project_Management.htm
Http://theoryofconstraints.blogspot.com/

For more details about the method,  some useful authors and experts in this domain whom you may want to read: Edward Deming, Peter Senge, Eliyahu Goldratt, Robert Newbold.