In previous posts, we have provided a brief introduction to the DMAIC cycle as a tool for improving processes. D. Define; M. Measure; A. Analyse; I. Improve and C. Control
In this post, we will be talking about the last stage – CONTROL.
“Apparently we´ve been able to improve by applying the previous stages, so the Process Control stage doesn´t matter so much… HUGE MISTAKE” “There´s nothing worse than seeing a rival overtake us right at the finish line”.
Throughout our career, We´ve seen excellent work carried out by teams that have managed to improve processes. After a while however, I´ve heard individuals say things like: “The process has fallen through again”; “The problems have come back”; “All that work we did… and it hasn´t been much use”
When this happens, we need to ask ourselves whether the control measures that we implemented in order for the process to inform us of its status were actually the best ones.
The purpose of the Control stage is to:
- Make sure that once the process has been analysed, corrected and stabilised (a good result is achieved and repeated), the conditions of this process are monitored via statistical techniques for process control.
Whenever we implement a new process or we correct an existing one and obtain better results than we had done previously, there are still emergent issues in the process. In other words, these problems are rooted in learning and occur when people who mainly “find it hard to change their habits” are implementing the process – a lack of familiarity with the new process can result in mistakes being made. It is important to take this into account during the implementation stage.
Once we have overcome the learning stage and any temporary difficulties, the process control should specify a series of statistical techniques for ensuring that the product has suitable characteristics by controlling the vital variables of the resulting product.
“A vital variable is something that can cause a deviation in the process if it is not controlled, thereby causing damage to the product or a deviation from its specifications”
- Alloy temperatures when soldering electronic circuit boards
- Dimensional values that could be affected by process variables
- The plastic pressure in the screw during the plastics extrusion process,
In previous posts, we always had to determine the vital and trivial variables of our product and/or process. The first things we have to do in order to devise the process control are:
- Establish tolerances for vital variables
- Keep vital variables within those tolerances by setting up suitable controls.
- Use measurement and control charts in order to monitor and control any variation in the vital variables.
How do we make sure that the vital variables are within their desired tolerances? There are basically two kinds of control method:
Throughout our career, We´ve noticed how we often we focus on behavioural methods, i.e. those that aim to inform workers of changes to their job, conduct or operations. Let´s just say that those methods are good… but they are not enough. We often hear people say things like:
“… I´ve changed the instruction, but the error has occurred again” “…it´s in the procedure but they haven´t read it” “… etc.
Error – Proof Methods.
“It´s good to do things well and do them first… but it´s even better to make it impossible to do things badly the first time”
Most of us have heard of famous mistake-proof methods (or Poka-Yoke, as they are known in Japanese). These systems are currently the best way of ensuring that processes are free of faults or errors because they eliminate repetitive tasks that depend on memory or being constantly alert, freeing up time for people to spend on tasks that have greater added value. Making a couple of mistakes is not acceptable… the goal is zero defects.
We must therefore make a decisive move towards a SIX SIGMA APPROACH:
Six Sigma approach
The CONTROL PLAN is a document that every organisation should have in order to monitor and control processes. The quality of the process and the product depend heavily on the quality and detail of this control plan.