Buffer   
Buffer Management  

Buffer 
If we apply the Five Focusing Steps, we can identify the constraint of our system and see a way to build the subordination mechanism. 
What we still need  to find is a way of protecting the constraint.  
The way to protect the constraint is to place a buffer in front of it, i.e. put ‘enough’ in front of it to keep it constantly occupied. 
We now have to ask ourselves two questions: 

1) ‘Enough’ what?  
2) How much is enough? 

1) Time. The unit of measurement for a buffer is time. In a production environment it is the amount of time we use to release material ‘early’ to  make the constraint work. In Project Management it is the amount of time we accumulate at the end of the project before the set delivery date. 

2) We consider a buffer to be made up of three zones,   zone one being the one nearest the constraint. 

What we now have to ask ourselves is: how much of this buffer can we use up in total safety? Only the external part, zone three. If we get into zone two we have to ask what’s happening and why what we were expecting hasn’t arrived. If we get into zone one, we have to take immediate action to get things moving. No ifs and buts, we must have things now. Otherwise we lose throughput. 

The length of the buffer is determined by the predictability, or interval of variation of the in-control processes that impact the constraint. The higher the variation, the longer the buffer time has to be. A perfectly in-control process could do without zone three. In the same way, in a completely out-of-control process,  zone three would be equivalent to zone one. So what should we do? 

It is essential to remember that we cannot predict the outcome of any managerial action if it is not taken in an environment made up of stable and in-control processes. Consequently, if we want to get the most from buffer management we have to make sure that: 

- all the major processes that impact the constraint are stable 
- all the major processes that impact the constraint have a manageable interval of variation. 
 

Buffer Management  

We know that the processes in our system are, by definition, affected by variation. As these processes are strongly interconnected, this variation inevitably influences the constraint’s performance and therefore the Throughput of the whole system. 
We have to understand that “an hour constraint time  is an hour of  system time”; we always have to make sure that the constraint works to the maximum of its capability in the time required. 
Buffer Management is a control mechanism that protects the constraint from variation in the processes that impact it. Buffer Management allows us to identify those areas that are not in control and to safeguard the constraint’s performance. 
Let’s consider what happens in a production environment in a company where we have applied the Five Focusing Steps and  developed a plan as a result.  In day-to-day reality, things will not always go the way we expect: what happens if  out-of-control processes interfere with this plan, or it gets held up due to behavior that is not in line with our scheduling?  
Buffer Management allows us to see when an order risks being delivered late before the situation gets out of hand and it’s too late to do anything about it. By means of this control mechanism, we can identify where and when an interruption in the production flow will occur. 

In this way, we reduce daily pressure and stress and we gain time to analyze and understand the causes that block the production flow. This analysis generates a positive feedback loop which strengthens the solution and creates an opportunity for further improvement. 

Buffer Management: Production 

In a manufacturing industry, the buffer controller visits the Finished Goods area and checks the number of products ready for shipping. Every unit under the scheduled number is considered missing and constitutes a ‘hole’ in the buffer. The way we react to a hole in the buffer depends on the time left before the delivery date. 

In zone three we are likely to find holes in the buffer, due to the transfer time  from the Raw Materials area to the Finished Goods area. 

In zone two we expect to see some units arrive in the Finished Goods area. The only reason for a unit to be missing in this area is because it’s been held up due to an interruption in the production flow.  
At this point the buffer controller checks up on where the missing units are. This way he gains knowledge of the problems the units encounter en route (e.g. large batches, machine breakdown, an absent worker, etc.) 

In zone one, we take immediate action to remedy the situation and guarantee on-time shipping. 
 

The normal trend for the arrival of units in the Finished Goods area can be illustrated by the following graph 
 

Thanks to the increased focus gained through Buffer Management, and  the preventive actions it allows us to take, we can expect the graph to change to the following 
 

What is the constraint in Project Management? 

The Critical Chain 

Exploiting (see  Five Focusing Steps) the Critical Chain means doing everything necessary to guarantee that tasks are not late according to the schedule, including detailed planning; protecting the duration of the entire project and not just the duration of single activities. 
In order to do this, we insert the project buffer which protects the conclusion date of the project from fluctuations along the Critical Chain. The length of this buffer is initially 50% of the length of the Critical Chain. Indeed, as we have already seen when discussing buffers, the size of the project buffer depends on the variation in the processes that impact the project. 

To protect ourselves from scarcity of resources and guarantee that the resources we have allocated along the Critical Chain are there when we need them, we use the resource buffer. 
There are two ways of activating the resource buffer mechanism: 
1) We send a warning message to resources that have to intervene along the Critical Chain to keep them ready to act (as we have seen, this works within a stable system) 
2) We allocate an alternative resource. This makes sense when tasks are subject to high variation. In this case, adding a resource means protecting the project completion date. 

Subordinating  (see Five Focusing Steps) the rest of the scheduling to the Critical Chain means making sure that all the other resources do whatever possible to guarantee that the task scheduling is respected. 
In this stage, we are faced with a conflict: start late/start early. In order to be a good project manager, we have to satisfy the client in terms of reduced lead time and so start the project activities early. 
On the other hand, in order to be a good project manager, we have to provide a realistic schedule, and so we have to start the various activities late. 
 

The way to get out of this conflict is to shift the beginning of activities as far ahead as possible while guaranteeing that the project will not finish late.  
The solution to the conflict lies in deciding to protect only those activities that lie along the Critical Chain. In order to do this, we insert another kind of buffer, called feeding buffer. This protects those activities that are not on the Critical Chain, but do impact it and influence the project completion time.