The perfect world!
In a perfect world, you would monitor and check every part, product or process against tolerances to ensure you are meeting the expectations of your customer, regulatory bodies and your own company standards. This would include all of the raw materials or sub-components that go into a product or process. This is not realistic in a lot of cases due to the sheer volume or speeds of most manufacturing processes.
So what do you do? Well, you would develop a testing procedure that collected data on the key variables that are important to meet customer expectations. How often do you test these parameters? Answer: How long can you leave it before it becomes a problem? That answer will upset a few people but it is a fact. We do not want to over test (too costly) or under test (defects, rejects, re-work, scrap), so the best way is to apply Statistical methods to determine the right time to sample for each situation.
Continuous Improvement – Consistent performance!
See if this sounds familiar. A list of current problems are developed (top 5 or 10) and they are typically related to some measurable such as downtime, defects, rejects, scrap, customer complaints – the big killers of efficiency. What do all of these items have in common? You failed to predict they were going to happen…you never saw them coming! Why? Because there was not a method in place to highlight the problem before it became a hit on efficiency. So you investigate and identify the problem – bad material, running too fast, temperature too low, pressure too high – root cause determination.
Next you take some action to correct the problem like make an adjustment or change materials but if the solution does not involve monitoring a cause on a regular timed basis, it will be back..tomorrow, next week or next month. I like to call this repetitive failure…you didn’t really fix the problem so it will be back! Soon, other items will be higher on the hit list and get the attention but the old items will be back unless you do something about them.
What is Continuous Improvement?
It really means learning from your mistakes and making sure they do not re-appear.
Identifying your mistakes involves a process called event analysis. This is the capturing of data like a defect or downtime event when it happens and then using the data collected to analyze, understand and eliminate the problem. The capturing of this data requires the capturing of two types of data for each event, Qualifying and Quantifying the event.
Quantify is how bad was it – Items like minutes down, number of defective cases or products and cost of the event. We all react to some measure of an event and treat it different accordingly. One downtime event could be 1 minute long or 2 hours long – you would treat them differently even though both happened once.
Qualify is just as important. These are what we call the symptoms of the event. What was happening at the time of the event. Information like product being made, team running the line, production line, size, colour, materials or customer, all of these items may be improtant for understanding and ultimately solving the problem.. For downtime it would include the root cause and items such as type or category like electrical, mechanical or operational to name a few. All of this information collectively and separately is valuable information for the investigative process and eliminating the root cause in the future.
A simple rule is if you did not collect it at the time of the event then you cannot analyze by it later. Think of the work police do at a crime scene, they collect everything they can so it can be used to solve a crime later.
Then what is Consistent Performance?
It is the processes you put in place to ensure losses or events do not re-occur once you have investigated them and determined the cause.
This would involve creating a monitoring process and using statistical tools like Specification, Control and Target limits and other tools to ensure you identify and react to change before it becomes a defect or downtime cause. You can determine how often you need to monitor a setting or measurement over time by reviewing the data on a timely basis like end of shift meetings. You will find that some parameters need to be monitored once an hour because of variation issues, where other more stable ones require monitoring once a shift or once a day.
The Quality Window product provides you with an adaptable tool that provides timely and cost effective solutions for both Event driven – continuous improvement, as well as, regular timed sampling for consistent performance applications. A key feature of the Quality Window product is its rapid change ability allowing you to keep pace with the fast pace of change on the factory floor.
Examples of these two types of applications are provided with Quality Window and can be found in the QW5/Samples directory in the Busitech folder on your computer.