Noise, vibration and harshness (NVH) testing on the factory floor poses its share of challenges....

3 Reasons Why Engineers Regret DIY Quality Monitoring

By Joe Ventimiglio, Sciemetric 

From Sciemetric's Signature newsletter. Subscribe now.

 

Every year, I meet with hundreds of smart manufacturing, process and quality engineers across the U.S. It’s no surprise that one of the most common ways to monitor the manufacturing processes on the line are the do-it-yourself options of rigging a PLC or building a system with a computer, acquisition card and coding software. These are appealing because it appears that one can meet the requirements at a lower cost. However, I have learned from these engineers that there are also downsides to the DIY approach, downsides that do not become apparent until it’s too late. 

 

1. It costs more than you think.

There is also the inherent risk of any software development project. The project has many cycles to get it right and often takes longer than planned, duplicating what is readily available on the market. The end result may work but it’s unproven and usually based on the knowledge and skill set of one individual. The software personality reflects that individual’s;  and only he or she knows it, making it a challenge to service it in the future.
The most reliable, fastest and easiest to use applications evolve over a long period of time based on user feedback; this is what an engineer taps into when buying an off-the-shelf system. Experience, and proven technology designed for the requirement.

Open a spreadsheet and do an honest accounting that includes the time to implement the system and see if the cost savings and potential risk of DIY are worth it.

 

2. With a PLC you get…a PLC

Using an analog card on a PLC is a cheap and easy option, but it really limits the kind of measurements you can do. Those measurements are only enough in a small number of cases; mostly the risk that those limited features are really capable of finding quality issues is greater. If data is collected, it is often bare bones: no signatures for real visibility into what occurred in the process. For backwards traceability – real traceability -- to intelligently fix process problems, you don’t have enough to go on. Which leads me to:

 

3. When there’s a problem, you’ve got…nothing.

Have you heard anything in the news lately about quality spills? First is the challenge of figuring out where the problem originated. Then what happened so that it can be addressed. Customers must be alerted and a recall undertaken and then determining who is affected is another issue. None of this can be done efficiently or in a way that protects the manufacturer’s reputation without having real data. Because it would all be guesswork. With data, you can get the facts.

That doesn’t just mean basic facts in a database somewhere related to a serial number, it means deeply detailed information on what happened with the assembly throughout every critical operation. It also means having the right tools, not just spreadsheets, to make sense of the data and turn it into information. So that instead of recalling thousands of units you could find the dozens that actually have the issue. Imagine how impressed your customers would be.

Though I get to say “I told you so” a lot, I take no pleasure in it. If you care about quality and if you care about traceability for each assembled part, look beyond DIY.

 

Joe is the director of sales, automotive/industrial at Sciemetric. He can be reached at joev@sciemetric.com.