Listen:

HxGN RadioPodcast

[PODCAST]: Vital Statistics: Driving Productivity with Statistical Process Control

As industrial manufacturers seek innovative new ways to improve productivity, product life cycle management is becoming an increasingly important factor in business success.  In today’s podcast, we explore how statistical data can provide the process visibility required to make efficiency savings across the supply chain and improve production outcomes. To listen to more episodes from HxGN Radio, visit our channels on iTunesSoundCloud or Stitcher.

Welcome to HxGN Radio. This is your host Bill Fetter. As industrial manufacturers seek innovative new ways to improve productivity, product life cycle management is becoming an increasingly important factor in business success. Our podcast today looks at how statistical data can provide the process visibility required to make efficiency savings across the supply chain and improve production outcomes. Statistical process control specialist, Q-DAS, was acquired by Hexagon in March 2015 and is part of the Hexagon Manufacturing Intelligence business.  In today’s episode we’re talking to Tom Stewart who is president of Q-DAS Inc., the North American subsidiary of Q-DAS.

BF: Tom, thanks for joining us today.

TS: You’re welcome Bill. Thank you.

BF: Alright Tom, let’s get right into it. For those of our listeners who may not be familiar with statistical process control, could you give us a 30,000-foot overview of what SPC is?

TS: Sure, I’d be happy to do that. Statistical process control is the matter of collecting data during the manufacturing process and determining whether the process is stable, in control, and trying to determine the probability that parts will be in spec or out of spec.

BF: OK, so those are the key benefits?

TS: Yes, they are.

BF: OK. So what are some of the unique challenges that manufacturers face when they’re trying to apply statistical processes to manufacturing systems?

TS: Well, the major issues are that in many cases the traditional methods of statistical process control, which were originally derived in the 50s, 60s and 70s are simply not applicable today. So in a given plant, while they’re making production parts they have to be concerned about tool wear, and tool life, and targeting a process, and so many process parameters.

BF: So you said some of those processes that were developed in the 60s and 70s and are no longer applicable today. Why is that exactly? Has the times changed that much or do we just have a different way of looking at things?

TS: Yeah. In those days the production process was more or less linear. For example, in the automotive industry – transfer line based, where a piece part would be processed in one operation and then sequentially go to the next and the next and the next. So in those days, they were taking the piece parts off of the production line in the order they were manufactured and then measuring them in the order that they were manufactured. And in that case they could develop a subgrouping of the data and also in those cases, something called the Western Electric Rules: runs, trends, middle thirds, were more applicable than they are today.

BF: OK. So how does today’s environment change the way that we should be looking at SPC then?

TS: Well in today’s environment the manufacturing systems tend to be more agile in nature. Also, the materials and the tooling, you know technology has greatly improved. So these days we’re measuring pieces of perhaps one piece from one machine a shift, and in that case you need enhanced statistical methods.

BF: So how does this all tie in with industrial metrology, specifically?

TS: It really ties in nicely. As the technology, also of the metrology systems has improved over the years. Their ability to include traceability, not only serial numbers but machines, fixtures, tools, feature groups. This information is now coming from the metrology systems and can be correlated in the Q-DAS product to make sense of the data that we’re receiving.

BF: OK. But that’s not the only place you might be getting data, is it? There’s other places that you might be gathering data to look at process?

TS: That’s correct. If you look at the typical manufacturing plant, they have a lot of suppliers. Suppliers are producing the same kind of part, shipping it to the plant. When those parts arrive they always have variability in them and so it’s important to understand which supplier the components are coming from and that can be recorded as well.

BF: So that’s part of the idea of the supply chain, SPC in the supply chain?

TS: That’s correct.

BF: OK. So what else should you be thinking about if you’re thinking about how do I apply SPC techniques and incorporating them into my entire supply chain?

TS: Well, it always depends upon the knowledge of the process. In our case we partner with the customers, we partner with the suppliers because they’re the people that understand their processes, how they’re producing their parts. And then we adapt the systems to suit: a) how the data’s coming, b) how it’s been recorded, and c) the reason it was recorded. For example, I would look at data differently if it recorded as a station check or tool change check, as if it were recorded for the standard production.

BF: OK. So this sounds awfully complicated. Is this something that everybody’s doing or is this something that only the biggest manufacturers are doing?

TS: We’re seeing it as a coming trend. It certainly started with the larger producers but it’s becoming more and more important for the middle production kinds of facilities and also­­ job shops.

BF: So how does a company that’s smaller and maybe not so familiar with these techniques, how do they get into this and how do they make sure they’re doing the right thing?

TS: Well, if they have the desire to support continuous improvement, then they have to record the measured values of the process and if they are working with a company like Q-DAS, we’re instructing them, we’re helping them, we’re training them, teaching them about what we do. Taking our knowledge and combining it with their process knowledge to come up with the products and support services that make it beneficial to them.

BF: OK, so it sounds like there’s a good component of consulting as well as software in what Q-DAS is offering here.

TS: Yeah. It’s very, very important. We have process experts. We have software experts. Actually, we have very, very good people around the world. But it also needs the customers’ input, the customers’ understanding of their processes, because they’re all different.

BF: OK, but the processes may be different but I understand a lot of the tools or the methodologies that you use are often out of the same toolbox.

TS: From us? Yeah. We use a standardised set of software tools that are adaptable to meet the customer’s needs.

BF: So you would work with the customer to figure out how their process applies to the available tool set?

TS: Absolutely. Absolutely.

BF: OK. So what kind of industries… I know Q-DAS has traditionally been involved in the automotive industry. What other kinds of industries could benefit from this kind of an approach to managing their quality process?

TS: Well although a large portion of our business is in the automotive industry we’re also very active and promoting the concepts inside the aircraft, heavy duty, energy markets. In fact, I’ve said a few times that aircraft is a lot more similar to automotive than a lot of people would think because of the exotic materials and tools.

BF: Are some of the processes the same then, is that how it carries through?

TS: Well for the aircraft industry it’s a little bit more important to understand the capability of the tool and the machine. They have to measure the characteristics because this is the customer deliverable but from a production standpoint they really care about the capability of the tools and the machines and we can do that as well.

BF: OK. Normal and non-normal data. In manufacturing, which one do you tend to get?

TS: Well we tend to get non-normal data…

BF: So what do you do about that?

TS: Well, inside of a Q-DAS product is a statistical engine that has the ability to determine the correct data model for any given set of data and because we normalise it at that point we’re able to compare non-normal data and normal data together  in the same setting.

BF: OK, so what does that buy you?

TS: That buys us clarity into the process that eliminates the confusion about somebody transforming the data and making an interpretation on the data. We let the data speak for itself.

BF: OK, and why is that a benefit going forward?

TS: Well the key benefit is that with other processes where you’re trying to manually determine the normality of the data, I might do it differently than you would do it, than someone else would do it. And then somebody else has to figure out six months from now how it was done. And with our system we’re using a configuration of evaluation which equates to a set of business rules for analysing the data that is customer specific.

BF: OK. So how would a customer go about building those business rules then? Is that something a customer normally would have?

TS: They frequently do, but we’re relying on the international norms and standards. ASQ, AIAG, the VDA… we’re just applying them in a very pragmatic and reasonable way.

BF: OK. How do you see the business of SPC developing over the next few years? Where do you think this is going?

TS: Well, with the advent of big data and Industry 4.0, there’s more and more demand to collect more and more data and from our perspective the need for the correct kinds of analysis will only increase. But also we’re taking it a step earlier in the process, determining the capability of the metrology systems and the manufacturing systems ahead of time to determine what kinds of data needs to be collected.

BF: So that’s sort of a planning process in setting up a system?

TS: Yeah.

BF: So you make sure what you’re designing in terms of a manufacturing system actually is going to tell you what you want it to tell you?

TS: That’s correct and that’s why we’re developing the relationship with the customer and the partnership with the customer to understand the process from their perspective, applying our tools to give them the information and analysis they need.

BF: OK, alright. So if you had to give someone who’s thinking about deploying SPC in their business… what are some of the most important things that they need to be thinking about? What’s the biggest benefit that they might accrue to applying these techniques into their business?

TS: It does depend upon the industry, but step one is understanding the capability of the metrology systems, understanding the capability of the manufacturing systems and then with their process knowledge defining what kind of data should be collected, how frequently it should be collected and then what you do about problems you perceive in the process.

BF: OK, and what’s the ultimate benefit then?

TS: Productivity and quality.

BF: Very good. Alright well thank you very much Tom, we appreciate your time today, thanks for being our guest.

To our listeners you can learn more about Hexagon Manufacturing Intelligence at hexagonmi.com. Tune in for more episodes from HxGN Radio on iTunes, SoundCloud, or Stitcher radio. Thanks for listening.