The critical chain is a tool developed by Goldratt to improve project management processes. Based on field observations and on the principles of the stress theories he developed, he devised an effective method that challenges the principles of classic planning.
One of the applications of TOC developed by Goldratt, concerns the improvement of project management. He found that many projects did not respect the elements initially defined: The deadline and the budget. According to him, for the reasons that the traditional planning method (method PERT) does not take into account either resource limitation or the fact that the same resource can be allocated to parallel tasks.
It will build on 3 principles to develop a new project management method called CCPM for Critical Chain Project management.
- 1er principle: in a classical system, each task is associated with an uncertainty period. Instead, Goldratt proposes to use this time limit at the global level.
- 2nd principle: The time limit for each task is given by the actual availability of resources and not by the infinite availability of these resources.
- 3nd principle: The belief that multitasking is more efficient than the Monotask is false. The same person should only be allocated to one task at a time to be effective and obtain a quality result.
Demonstrating the superiority of single task mode to multitasking mode
Multitasking is a mode of operation where the same person simultaneously advances different tasks.
The only task is to start a task only if the previous one is finished.
What we recommend Management by the critical chain is to work in a monotasking. Here’s why.
1-Imagine that we have to do 3 tasks and each takes 3 days. In monotasking, we get the following schedule:
Let’s set up our performance indicator D which will consist of measuring the average time to accomplish the tasks. In our case, this one is: (3 + 6 + 9)/3 = 6 days. In other words, on average, our tasks are over in 6 days.
2-Now let’s say we decide to run in multitasking mode and start a task when the previous task is only 1/3 advanced. We get the following schedule:
3-by generalizing, we obtain the following formulas:
In monotasking: D = (1 + s) * n * (n + 1)/2
Multitasking: d ‘ = (n2 -(n-1)/2) + S * (n3 -N (n-1)/2)
For example, for a project of 15 tasks with 10% Setup, a/d ratio of 4.13, a total delay of 4.13 times greater in multitasks than in one task.
- S: delay in percentage of Setup (change time required to start or resume the new task)
- N: The number of tasks to perform.
It is pointless to try to do several things at once because in the end it takes us more time. You have to stay focused on your work. That is not to say that we should never do 2 things at once, but that it must meet certain rules:
- It is easy to move from one to the other.
- We cannot move forward in the immediate future in this task because we are waiting for a few things.
It should be noted that in addition, we do not take into account the level of fatigue that results from the fact that we have to “think about everything“.
Multitasking: A real public health problem
Many studies * demonstrate the adverse effects of multitasking. We are now invaded by small tasks that range from reading an SMS, to writing an email…
All these little tasks are “exciting“: It gives us the impression of moving forward and generates dopamine in the brain.
This research shows that it is not without consequence and speaks of a “Cognitive cost“. These small daily tasks actually generate a significant decrease in IQ and a decrease in brain density. As well as people not sleeping, or smoking drugs.
1-Establish the list of tasks
The first step is to make the list of the different tasks to carry out the project. As far as possible, this listing must be exhaustive.
2-Allocate a time
For each task, it is estimated that there is a duration and identifies the necessary human and material resources. For the estimation of durations, 2 methods oppose:
- Goldratt proposes to use the median of the times used generally for this type of task.
- The Product Development Institute1 proposes to use the averagetime used generally for this type of task.
3-Defining the critical path
From the desired end of the project to go back to the beginning by having the tasks with a delay ” at the latest “. Make the diagram taking into account the fact that tasks can be performed in parallel or not, and define the necessary background.
At this stage, resources are still considered as infinite (traditional method).
4 – Divide the time of each task by 2
When preparing a schedule, the tendency is to overestimate the duration. This is for very different reasons: bad experiences, to protect itself from the fact that management will want to reduce the time…
For all that, and despite these protections, most projects end up late. The reasons proposed by Goldratt are as follows:
- student syndrome: Every person will have a tendency to procrastinate and will take it at the last moment to do an action when he could have done it beforehand.
- Interdependence: Tasks are often linked to each other, if one takes a delay, the others too.
- Parkinson’s Law2 : Any person will tend to use the entire time allotted, regardless of the duration of the term.
- Carlson’s Law: Continuous work takes less time and energy than when it is done in several times.
- the Logic of Honor: A professional will have a heart to provide a psychologically bounded indication by not indicating too short a time (not likely or impossible to hold), not indicating too long a period (doubt about its competence).
The law of Distribution of project deadlines
It is admitted that the duration of the projects follows a beta distribution law. This law tells us that if we want to increase our probability of being on time by 30 to 40%, we actually need to double our time to estimate the duration of each task. In other words, we are very unproductive and poorly performing.
5-Identify and manage resource conflicts
For each task, we will go into the details of the resources allocated and define whether conflicts are present, that is, whether the same resource is allocated to 2 tasks in parallel.
Always starting from the end of the project, we will manage the different conflicts. It will be possible depending on the feasibility:
- Re-arrange the tasks by going back to the project.
- Re-allocate tasks to different resources.
6-Identify the critical chain
Once conflict management is done, the critical chain will be calculated. This is equal to the time set by the longest task chain.
7 – Add an overall buffet half the time truncated
In the logic of Goldratt, we have not yet taken into account the potential delays that the projects can take. The method then proposes to set up a global buffer, “project Buffer“, which is equal to the time of the critical chain divided by 2. This principle is found in the planning of the construction sector under the term ” time frame “3.
” feeding buffers ” are also placed for each task that is not in the critical chain. The calculation of these buffers is on the same principle as the global buffer.
8-Set up the ” alarms “
The method also proposes to anticipate tasks by preventing resources. Clearly, the project manager will have the ” alarms ” 1 week before the start of each task, an alarm that will allow him to prevent the resource in charge of the task that she will soon have to allocate time for the task in question.
9-Piloting the project
Once all this planning has been put in place, the project manager will only have to follow the various alarms and make sure that the resources are well available.
Beyond that, the method proposes to use the Fever Chart (name given to the temperature curves that were put at the foot of a patient’s bed). The Fever Chart proposes to use the reading of the consumption of the buffers to drive the projects. We find:
- in X: Progress of the project (s).
- in Y: the level of consumption of the project buffers.
- Green Zone : It shows a final consumption of not more than 75% of the buffers. As long as the project or projects are in this area, there is no need to act.
- Orange Zone : It shows a final consumption of 75 and 100% of the buffers. If a project or actions are in this area, then you have to understand why and take the necessary steps to get back into the Green Zone.
- Red Zone : It shows a final consumption of more than 100% of the buffers, in other words we are asleep and already late. We need to set up a crisis unit to get back into the Green Zone. It should be noted that it is not advisable to reschedule tasks too frequently4.
1 – Product Development Institute (1999) – Tutorial: Goldratt’s Critical Chain method
2 – C. N. Parkinson (1957) – Parkinson Law
3 – R. H. Clough, G. A. Sears (1991) – Construction project Management
4 – R. V. Newbold (1998) – Project Management in the fast lane, applying the theory of constraint
E. M. Goldratt (1997) – Critical Chain
W. Herroelen, R. Leus (2001) – On the merits and pitfalls of critical chain scheduling
P. M. Chawan, G. P. Gaikwad, P. S. Gosavi (2012) – CCPM : TOC based project management technique