Planning for the management of constraints under a dynamic environment of uncertainty and optimisation of scarce resources – by Lewis Trigger
Download the full article – click here!
- Effectively manage uncertainty and maximise the use of scarce resources in a dynamic project environment.
- Shift your thinking from localised efficiency towards system effectiveness.
Projects by their very nature are plagued by high levels of uncertainty. The single, most outstanding characteristic of a project environment is uncertainty. Planning and managing uncertainty is critical for project success.
Another dominant attribute of most projects is the scarcity of critical resources. Resource constraints are particularly prominent to a dynamic multi project environment.
The traditional approach to Project management at time falters under the stress resulting from the high level of uncertainty together with constrained resources inherent to a project environment. A new approach is needed to deal with the mounting technical and management risks placed upon scarce resources in the development and deployment of sophisticated products, systems and facilities. This need for a new and effective approach is found in the contemporary body of knowledge known as the Theory of Constraints (TOC).
TOC provides a common sense workable solution to the age old management challenge of providing more with what we have. Within a project context what this means is focusing on the true critical paths of projects with a totally fresh innovative approach to safety buffers and scarce resource management without necessarily having to compromise “time to market”, performance, and cost. TOC goes on to further provide practical effective solutions to dissolving the complexity of a dynamic multi project environment via an inventive scheduling methodology that enables an effective staggering of project time schedules in accordance with the system constraints.
TOC is generally best known through the management bestseller book “The Goal” by Dr. Eli Goldratt and his project management book “The Critical Chain”.
Goldratt‟s innovative project management approach has been adopted as a leading part of the “arsenal” of managerial techniques by a wide variety of successful organizations around the world, including Boeing, Lucent Technologies, Delta Airlines, Israel Aircraft Industries, Daimler Chrysler, Westinghouse, The United States United States Air Force, Navy and Marines, The Israeli Defence Forces; and more recently NASA. TOC’s methodology has
also begun making inroads in Australia with prominent companies such as BHP.
TOC’s Building Blocks:
In order to have an understanding of this innovative approach it is necessary to be introduced to TOC’s building blocks. The methodology of TOC’s building blocks is broken down into 5 steps:
- Identifying the constraint (i.e. the “Bottleneck”)
- Exploit the Constraint (i.e. “Take what you’ve got to get 100% use of the constraint”)
- Subordinate to the Constraint (i.e. “Rearrange others to what you need to maintain 100% of the constraint”)
- Elevate the Constraint (i.e. “Innovate to get greater than 100%)
- Identify the new Constraint – continual improvement loop.
The above 5 steps provide the key to effectively managing uncertainty and to maximizing usage of scarce resources. Note that the 5 steps all focus on the System constraint.
Why focus on the constraint? Let’s think for a moment by going back to basics and asking ourselves the question “what really is the ultimate goal of commercial organizations?” In the final analysis the answer is obviously profit. How much profit? As much as possible! What we are talking about here is an infinite goal, as opposed to a commercial target which is obtainable and as such finite. Since we are talking about something infinite, we can logically conclude that there must be a constraint in the system, otherwise we would reach it. Goldratt, the founder of TOC, who is a Dr. of Physics, has concluded that if we wish to improve the performance of a system, meaning to bring it closer to its goal, then the area to focus on is the constraint itself. By managing the constraint we can gain the most leverage in improving the performance of our organization.
Is the idea of the pursuit of an infinite Goal equally valid for a non commercial organization? Take for example government agencies of defence, health, education ….. If so does the logic of focusing on their System Constraints in order to gain the most leverage in improving performance make sense?
How best then to manage the constraint? TOC has developed an extremely powerful methodology whose strength is in its simplicity and common sense approach to problem solving. At the heart of this methodology are 5 sequential steps:
The first step is to identify the System Constraint. Identifying the correct “entity” is critical, since the whole doctrine is based on managing the constraint. At times the answer is obvious, as is the case of some capacity bottlenecks. At other times the constraint can be more subtle, such as incorrect performance
measurements. Furthermore one must take care in distinguishing between problems that are symptoms of incorrectly managing the constraint but not in themselves constraints.
The second step is to “exploit” the constraint, or in other words how can we achieve close to 100% utilization of the constraint. Here TOC differs from traditional thinking. Under more traditional thinking, when confronted with a bottleneck, the thing to do is to “break it”. However TOC reminds us that normally we are required to invest additional capital and resources in “breaking” a constraint. Before we go down that road then more than likely the constraint or bottleneck provides us with additional capacity that we haven’t used. For example, if we have a work centre that is a true bottleneck, before investing in additional capacity for it, what we should do is to ensure that we don’t waste the capacity it has. How? – through many steps; for example by having it work more shifts; by placing a safety buffer of work material in front of it; by applying quality control before the bottleneck so it as not to waste its capacity on damaged work in process; and the list of common sense intuitive solutions flows on.
The third sequential step is to “subordinate” non constraint factors to the constraint. What this means is that non constraints have to support the constraint by being available to either feed the constraint or NOT to “interfere” with the constraint’s operation hence reducing the constraint’s throughput. That availability is made possible by gating the entry of work assignments in accordance to the pace of the constraint. Resource centres need to work in accordance to the pace of the constraint rather than aiming to work according to their own maximum utilization. We need to remember that ultimately the throughput of the entire system is determined by the throughput of the constraint. Not only is no benefit gained by pushing more work assignments into the system beyond the pace of constraint; but by pushing more work into the system we inadvertently create disorder and prevent non constraints from being available for the constraint.
The fourth step is referred to as “elevating” the constraint. It is at this stage that we go to work redesigning the way we would normally operate the constraint, so as to gain additional throughput from the constraint itself. Again remember that the overall performance of our system is dependent on the performance of the constraint. If for example we can alleviate some of the workload on the constraint by off loading to non constraints, which by virtue of being non constraints have additional capacity, then we have increased the overall performance of the system. There may come a point where we break the constraint by investing additional resources; however this will involve doing a cost benefit analysis.
Once we reach the point of breaking the constraint it means that invariably we have come to a new constraint. Remember, by virtue of the fact that we are moving towards an infinite goal, we must then by logic come up against a new constraint. Hence, we have now come around in an upward spiral progression and reached the fifth step where we need to identify the new constraint and once more go through the sequential steps that lead us on the path of continual improvement.
Video – required viewing
View the movie “The Goal” – 45 minutes
During the viewing jot down the practical steps taken by Alex Rogo in his plant that reflect:
- “exploiting” the constraint
- “subordinating” to the constraint
- “elevating” the constraint
It is important to bear in mind that the story of Alex Rogo’s plant is an analogy and in fact all working environments are a “plant” of some kind where assignments need to be completed in a reality of uncertainty and within constraints and resources need to be effectively synchronised. The fundamental TOC building blocks used by Alex Rogo are universally applicable, and can be equally used with the same positive impact to a project environment.
Goldratt, E.M., Cox, J., 2004. The Goal.
Read pages 94 through to 119.
Answer then following questions from the “Herbie” story:
- What is the fundamental characteristic that defined “Herbie” as the system constraint?
- By placing Herbie at the front of the line of the scouts how did Alex Rogo evoke the principle of “exploiting” the constraint?
- How was the principle of “subordinating the non constraints to the constraint” evoked and why was it critical to the success of the hike?
- How was the principle of “elevating the constraint” evoked?
- What is connection between the “dice game” and resource management?
TOC & Project Management
The Traditional Approach to Safety Buffers:
Traditionally safety buffers are applied to individual tasks. By adding a time safety buffer, what we are intuitively doing is estimating the time required that will allow us a high probability of completing that task on time. In terms of statistics, given the dominant characteristic of uncertainty that is inherent to project tasks, the probability distribution function of most tasks is very much skewed in the direction of the higher probability numbers. In “layman” terms what that means is that at times estimation with an 80% – 90% probability of success is 3 to 4 times higher than the average time estimation.
Goldratt, E.M., (1997). Critical Chain. Pages 40 – 46.
An experienced project manager will want to add safety buffers to the individual tasks, especially where his / her integrity is at stake. Invariably, the need to add safety buffers leads to a conflict between the project manager and top management who have typically committed the organization to a time schedule that is short on its safety buffers. Remember that in most cases where there is a competitive environment, in order for an organization to be awarded a project, it has had to commit to a “tight” time schedule that has had to ignore the demands for safety buffers (otherwise the organization / project team would not be awarded the project.)
Think back to a project you have managed. As the Project Manager, what was your “gut” feeling towards the official schedule, scope, and budget that was announced with the kick – off stage of your project? Was there a sense of conflict between your personal integrity as professional and the official contractual agreement that the organization has committed itself to?
The above inherent “conflict” is paradoxically the paradigm that Eli Goldratt uses in developing his innovation solution. TOC claims that even where task managers and project managers succeed in planning in significant safety buffers, for the vast majority of cases these safety buffers do not provide us with the protection we anticipated. In fact TOC claims that the very use of task safety buffers exacerbates the problem. A point in case is the “Student Syndrome”. By requesting more time to complete a task we invariably put off doing the task until the very last moment. By delaying commencement we have compromised the chances of us finishing on time. Should “Murphy” hit us (as he invariably does) we find ourselves exposed with no protection and the task ends up running over schedule.
Goldratt, E.M., (1997). Critical Chain. Pages 118 (commence at bottom of page “Then I ask….”) – 128.
Jot down from your personal experience an example that illustrates each of the principles elaborated on in the reading. An example for each principle. Note that the principles outlines by Goldratt illuminate how safety buffers fail to provide us with the protection we expected. More than that though, why is
it that this mechanism of task safety buffers can exacerbate the problem and in end exposes us to a higher probability of not completing the task on time?
Identifying the Project Constraint:
TOC’s solution to a project environment is based on the TOC 5 building blocks earlier elaborated on. The first step is to “Identify the Constraint”. If you were to ask any experienced project manager where the constraint in your project is, then probably he / she would answer that the constraint is constantly moving. For example he / she could say “a month ago it was when we were up to ears with problems of integration of the laboratory proto type; today we are bogged down in testing; and next month I foresee the obstacles awaiting us as we launch into the beta site stage”. What in fact the project manager is describing is sequential segments of the critical path. We can now conclude that the constraint of any single project is its critical path.
In most cases, project managers identify the critical path by use of the PERT or CPM logic. According to PERT / CPM, the critical path is the longest sequential path of tasks based on a logical dependency and early start. By logical dependency we mean for example that activity “C” cannot commence until activity “B” has been completed. Goldratt reminds us however, that the “true” critical path needs to consider not only the logical dependency but also the logistical dependency between tasks. By logistical dependency what we mean is if any two tasks or more require at the same time the same constrained resource, be it manpower, tooling, workplace, etc. By taking into account the logistical dependency, in addition to the logical dependency, we have in fact uncovered the true critical path that Goldratt coins the “Critical Chain”.
Exploiting the Constraint:
Exploiting the constraint is the second step within the TOC logic. Primarily what that means is that we don’t want to waste the constraint, rather we want to do all we can to squeeze the maximum out of it. In the context of a single project this means that we need to “squeeze” the maximum out of the safety buffers that have been inherently build into the tasks on the critical path. TOC solution is simple but extremely powerful and innovative. Instead of adding on the safety buffers to the individual tasks, which itself causes problems, TOC takes the individual safety buffers and reallocates them to the end of the project into an aggregated safety buffer, called a Project Buffer. The project’s “critical chain” includes the Project Buffer. The “thumb rule” is to leave 50% of the original time estimation. From the amount that is deducted from the original time estimate, approximately 50% is aggregated into the Project Buffer. Mathematically the result is an overall project schedule whose true critical path, that includes the Project Buffer, is approximately 75% of the length of the original project schedule. Statistically, through a change of project team behaviour, we are able to increase the probability that we will in fact realistically succeed in completing our project in time, without compromising product specifications or budget. . The next extract from your text details the logic of how to “exploit the constraint” and he implementation mechanism.
Goldratt, E.M., (1997). Critical Chain. Pages 151 – 160.
The key to changing project team behaviour is in performance measurement. Project task teams are no longer measured according to the success of the individual tasks but rather according to the success of the entire project. This change in measurement encourages task members to both complete their task ASAP as well as to begin ASAP in the event that the preceding task finishes before its estimated time. Furthermore, providing the task team has done its very best yet “Murphy” has hit hard, it is legitimate for a task team to run over its schedule. We need to remember that this is the reason why we have a Project Buffer.
What would happen to the performance of a football team if the prime measurement for each of the team players was the number of goals each team member succeeded in scoring, rather than the number of goals the entire team succeeded in scoring? What can we learn from this re the performance of project teams?
Management needs to have a clear understanding of the importance of incorporating the Project Buffer into the project schedule. The entire logic of “exploiting constraint” requires the inclusion of the Project Buffer in the schedule. Where management does not understand this and decides to dispense with the Project Buffer, the result is a “death sentence” to both the project schedule and subsequent attempts to implement the TOC strategy.
Subordinating non constraints to the constraint:
In the context of the single project what this means is that all tasks and resources are subordinate in “serving” as well as “not disrupting” the “Critical Chain”. To this end TOC introduces the concept of the Feeding Buffer and Resource Buffer.
The Feeding Buffer:
Under traditional project management, it often occurs that a non critical path‟s turn into critical path as a result of the original non critical path’s tasks running over schedule to the extent that they over run the non critical path’s “slack time”.
TOC’s overall strategy is to do all that is logically possible so as not to endanger the project’s true critical path i.e. the “critical path”. To this end TOC introduces the concept of the Feeding Buffer. The Feeding Buffer uses the same logic as the Project Buffer except that it is applied for non critical paths. Refer back to pages 158 and 159 from the early text readings that elaborated on the Feeding Buffer.
The Resource Buffer:
Whenever a new resource is planned for a task on the “critical chain”, we need to evaluate the effect to the overall schedule should this resource not be available. A definite strategy that minimises the risk of non availability of such a resource is for us to plan for it to arrive earlier than when we actually need it. This strategy is known as the Resource Buffer. The advantage is twofold. Firstly it decreases the probability that a resource that we need won’t arrive in time. Should “Murphy” strike we have allowed enough time to “recuperate”.
Secondly it allows us to exploit an opportunity where a proceeding task completes its assignments earlier than expected, hence we are now able to commence earlier since we have the resources available to commence the present task.
Naturally there is a price to pay by planning to have resources made available earlier than required. This price needs to be weighed up against the risk of not having a critical resource available on time and the implication of that for the overall project. Generally speaking, the use of a Resource Buffer becomes more critical towards the latter stages of a project in which we have already consumed a large portion of our Project Buffer.
Vendors and Subcontractors:
Many projects are also dependent on vendors and subcontractors. Typically negotiations will be centred on price. TOC on the other hand emphasises “lead time” as being as important, if not more important, than price. If as a result of long lead times the project completion is delayed, the savings achieved through price negotiations can be dwarfed by the lost benefits from not completing the project. The lead times stated by vendors and subcontractors have their own safety buffers built in. TOC offers a logic that will allow suppliers to be more amenable to shorter lead times. By providing enough warning to the suppliers together with the necessary “kit” if appropriate (be it documentation, a partially completed product to be worked on by the contractor, etc …), the supplier can more easily commit himself to shorter lead times.
Paradoxically though, too much warning will boomerang back resulting in inflated lead times. A requested starting date that is too far out in the future will in turn push the supply to provide a lead time that is heavily padded with a large safety buffer. The further out in time the contractor is required to predict in terms of his availability, the more uncertain the supplier is in being able to commit to a tight schedule. Refer to the following extract from your textbook in order to illuminate further how TOC effectively integrates Vendors and Subcontractors.
Goldratt, E.M., (1997). Critical Chain. Pages 173 – 186.
A Dynamic Multi Project Environment:
Up till now we have been discussing the application of TOC on a single project. In reality we recognize that most projects are managed within a dynamic multi project environment in which project tasks are constantly competing for shared resources. The logic used by TOC for a single project certainly assists in protecting the project from evils of Murphy, but in itself it is not enough to allow the project to weather the storm of a dynamic multi project environment in which there is constant resource contention.
TOC evokes the same logic that it has developed for synchronizing resources in a dynamic job shop1 scenario. The key to success is to limit the number of
1 A production floor that is characterized by shared resources that need to be constantly synchronized in order to full fill a dynamic product mix. Typically we see the production of small batches of a large variety of different products with each product type batch requiring a different sequence of processing steps and a change of set up for the individual work centers.
projects in execution based on the most limited resources. Logically an organization can only complete within a certain time frame as many projects as it can within the constraints of its most limited resources, i.e. the system constraints. Releasing more projects than can be processed by the system constraints will only spread critical shared resources thin and inevitably will fall into the trap of bad multi tasking. Enforce this rule even if it means leaving some resources “unproductive” part of the time. Remember that what is important is to be effective and not necessarily efficient. One way of enforcing this rule is to adopt a process of „pipelining‟. Pipelining releases work in a metered fashion, similar to traffic lights releasing cars into the freeway. By concentrating resources on fewer projects, projects are completed faster; capacity opens up; and the organization’s throughput of projects actually goes up.
A side benefit of Pipelining is that the extra time (before projects can be released into execution) can be used for preparation or “full-kitting.” At last, the organization has ample time to properly prepare drawings and specifications, get approvals, order materials, and test plans.
Paradoxically, by limiting the number of projects in the pipeline at any one time we in effect increase our productivity and complete more projects in a shorter time, within specs and within cost. The following text extract elaborates on the technique of TOC to effectively manage a dynamic project environment.
Goldratt, E.M., (1997). Critical Chain. Pages 230 – 237 (first paragraph)
View the DVD of Goldratt’s presentation on the Critical Chain (GSP Satellite series – 3 hours) for an excellent overview and summary of TOC’s implementation to a project environment.
Based on the readings assigned and the viewing of the 2 DVDs, we will go through a series of exercises and role playing so as to internalise the learning. In order to “exploit” our “constraint” of limited class time it is critical that you come to class prepared via the study guide.
Case Studies (recommended reading):
The two cases referred to are excellent examples of the successful implementation of TOC “Critical Chain” approach.
http://www.goldratt.com/f22.htm The F-22 is claimed by multiple sources to be the world‟s most effective air superiority fighter. The US Air Force states that the F-22 cannot be matched by any known or projected fighter aircraft. Air Chief Marshal Angus Houston, Chief of the Australian Defence Force, said in 2004 that the “F-22 will be the most outstanding fighter plane ever built.” Delta Air Lines: http://www.realization.com/pdf/APICSEXTRA_Nov_30_07.pdf
Revisit the Wembley Stadium project (GSN510 Blackboard site). Identify points in the critical path of that project where the implementation of TOC may have had an influence on the project outcome. Post your views to the discussion forum and read/comment on what other students have also identified.