This website uses cookies primarily for visitor analytics. Certain pages will ask you to fill in contact details to receive additional information. On these pages you have the option of having the site log your details for future visits. Indicating you want the site to remember your details will place a cookie on your device. To view our full cookie policy, please click here. You can also view it at any time by going to our Contact Us page.

Make machine vision work for you

22 September 2010

If machine vision is not implemented correctly it can be perceived as being worse than simply ignoring automated quality control. Mark Williamson, director of Stemmer Imaging and chairman of the UK Industrial Vision Association argues the case for intelligent implementation of what can be very effective tools.

When used effectively machine vision can competently weed out discrepancies in process and manufacture
When used effectively machine vision can competently weed out discrepancies in process and manufacture

Machine vision is often seen as an inspection bolt-on to ensure that final product specifications are met. Suppliers of supermarkets, for example, are often forced into inspecting their products to meet demanding specifications driven by a central QA programme, or by an historical lapse in product quality. Sometimes inspection systems are required purely to help justify the price differentiation on a premium version of a product.

I’ve seen many inspection vision systems turned off, as they do not always improve manufacturing quality, but can be perceived to increase waste and consequently cost. Specifications can be so tight that it is probable that the end user would never notice the majority of rejects, suggesting the irrelevancy of some of the demands placed on manufacturers.

Many people realise, after using machine vision for inspection in this way, that the specifications agreed are just unrealistic, or, that the inherent problem is in the manufacturing process. In this case all the vision system can do is confirm that there is a problem without either identifying or resolving the root cause.

In just a few days after installing such a system, the fact the line is just not going to deliver the required yields to specification becomes clear, and the vision system is ignored in favour of viable production.

It could be said that if the production line process control is good you just don’t need a vision inspection system at the end of a production line. Used correctly as part of the control system, in conjunction with realistic specifications that focus on checking features with significant cost, safety or reliability issues, can lead to real and significant benefits.

Integrating vision as part of the process control system, and using high speed cameras with analysis tools to visualise how the high speed processes vary during the process design itself, can enable process control to compensate for manufacturing variations, and can remove drift which left uncorrected could cause problems later on in production.

As a machine vision supplier we are seeing more opportunities and uses for machine vision both in the validation of production line design, as well as forming an integral part of the control system tuning a production line as it runs.

One hurdle remaining is to convince customers that there IS a time and place for specifications, but that these should be used judiciously. Failure to do this can make control and inspection a sham. It does not have to be this way.


Contact Details and Archive...

Related Articles...

Most Viewed Articles...

Print this page | E-mail this page