Quality in manufacturing

When managing a mass production line, our primary goal is to ensure on time delivery of products at the required quality level.

Your production process is made up of a long chain of separate and probably complex activities.

Quality problems will happen and although you are going to do everything possible to minimize them, it is something that you are always going to have to deal with.

The challenge of ensuring high quality of manufacturing increases when the manufacturing of your products are been done in other locations of your company or by a subcontractor. In both cases, your control over quality of process is limited.

Let us review the 8 most important rules that you need to have in mind when establishing a monitoring system for the quality of your production line:

If you choose a contract manufacturer to run your mass production, you can’t expect the same level of attention to your product as you would give it yourself.

Your hope is that your contract manufacturer provides you with required information about the manufacturing process and the Quality in Manufacturing during production.

But! You cannot always rely on receiving this important and complete information about the production process and product quality. And you cannot rely on receiving the information in a timely manner when you really need it.

Therefore, it is essential that you establish an independent and unbiased monitoring system. This is critical! You need to have complete control over the content and timing of the information coming from the manufacturing line.

Each stage of the production flow should be tested before your product can move on to the next.

Ideally you should deploy test stations (manual or automatic) throughout the production process. Testing should start with the inspection of incoming raw material, right through to the final stage prior to the delivery of finished goods to your clients.

Analyzing the testing data stored in all testing station located on your manufacturing line is very important.

This data provides invaluable information. It allows us to conduct root-cause analysis of quality issues, and over time improve production quality.

Production downtime is a nightmare for manufacturers. It may lead to significant delivery delays for your customers and damage the heart of your business.

It is essential that you have direct instant access to the data collected from the testing stations located on your manufacturing line.

You need to be able to react rapidly. Run root cause analysis. Identify and fix the problem, and resume full production as soon as possible.

Usually testing stations located on your manufacturing line (manual as well as automatic) measure several technical parameters.

The testing concludes with an indication – “Pass” or “Fail”. If the test result shows a “Pass”, then the unit is moved on to the next manufacturing stage. If the test result shows “Fail”, then the unit is sent to a technician for further analysis.

Why do we usually pay attention only to the “Pass” or “Fail” criteria? Why aren’t we interested in the other tested parameters?

The reason is information overload. When running a mass manufacturing line it is impossible to routinely “digest” all the detailed information collected from testing stations. We usually analyze this data in detail only when a quality problem is found and we are busy finding the root cause of the problem. If we get a ‘Pass’, then all this detailed information is typically forgotten.

A simple “Pass” or “Fail” gives you little or no information about edge cases – where one or more of the unit’s technical parameters is just within its allowed tolerance. “Edge” cases may lead to unit failure during operation, for example in extreme environments (cold, heat, humidity, electrical overload, impact etc.). For accurate and useful quality data analysis, you need to find a method that will let you routinely review and analyze the entire test data for the unit and analyze it in a meaningful way with other tested units, other testing stations and with historic test data.

The manufacturing process is a chain of separate but dependent assembly and testing processes, which together build your final product.

A technical problem created in one stage of the manufacturing process may only be identified in later test stage. For example, a defective button assembled on a unit may only be found during functional testing, several stages later.

You should expect test results from any of your manufacturing stages to potentially influence other stages in the process. Reviewing and analyzing the data collected in one testing station in isolation is just not sufficient.

In order to see the entire picture you need to collect and analyze the end to end results according to the severity and the frequency of each problem found.

The manufacturing of your products may be taking place on another continent. It may be taking place in the next room.

Either way you need to be alerted so that even if you aren’t there to keep a close eye on each stage in the process, you will still be aware of major problems the instant they happen. An automated alert mechanism that generates notifications about critical problems on the production line is an absolute must.

“A clever person troubleshoots the problems that a wise person avoids in the first place”.

Let’s be wise and fix quality issues before they happen. A good way to achieve this is to set up a predictive mechanism that analyzes trends within the testing results and alerts us to potential quality issues.