News Stay informed about the latest enterprise technology news and product updates.

True challenge for manufacturing QA comes at supply chain level

In this interview, Simon Jacobson of AMR Research explains how manufacturing IT managers should proceed in order to improve quality at their manufacturing firms.

The following interview on the subject of manufacturing quality improvement and IT technology took place recently between Simon Jacobson, a senior research analyst with Boston-based manufacturing consulting firm AMR Research, and Brenda Cole, assistant site editor of

Simon Jacobson is responsible for AMR's coverage of manufacturing operations and ERP providers for the midmarket; he also covers manufacturing compliance and quality management, manufacturing services organizations and MES providers. Quality improvement inside manufacturing plants has historically revolved around processes. How much of the responsibility for quality improvement can be shifted onto the IT team/IT manager?

Jacobson: When we look at improving quality management, and the role that IT has to play here, it's simple. In the past, we -- we being manufacturing organizations -- have always viewed IT as a purveyor of hope, that the newest solution will give us everything to solve our quality woes. IT enables and provides this ability and solutions to help, but the bigger piece is that IT has to be driven by the business.

If we really want to solve quality problems, modern software applications will help. They'll give us visibility into process performance, they'll give us a way to electronically digitize some of the work instructions or work processes around audits, around corrective actions and around process management. But ultimately, anything that's going to be improved in quality has to be an executive-level-driven process change. How is software being used to improve QA on the manufacturing plant floor and the back office?

Jacobson: That depends largely on the style of manufacturer. There's a host of different software applications that can either add to or detract from quality. What I mean by that is we can often start with the foundational aspect of process control, where we'll use a statistical process control application to provide real detailed information around how a certain process or part or material or component handles itself during the different stages of production. In some cases, you can also capture that level of data in a historian.

The big thing companies need to look at, if they want to apply IT or any software package in general to solving quality management issues, is the level of the problems they're trying to solve.

But the big thing companies need to look at, if they want to apply IT or any software package in general to solving quality management issues, is the level of the problems they're trying to solve. What I just described with the process control application is really a Level 1 issue, where you look at quality management within the individual functional domain or individual processes within a site.

The next step is Level 2, where companies have often used an MES application to bind together multiple pieces of quality processes across the entire manufacturing site. At this level, they take into account things like corrective action workflows. Things like audits, SOPs and documentation also come into play, as well as some of the test and inspection pieces. What happens at the next level?

Jacobson: The real challenge, where companies start to struggle or have issues, is when they want to expand up to Level 3 -- the enterprise or supply chain level. This is where you start to handle quality across multiple sites and you start to drive the same inside-the-four-walls competencies in very repeatable fashion.

So if I have nonconformance on site A, I'll have automatic triggers or at least a business process in place to drive an inspection for a similar nonconformance at Site B, because they're making the same product. This Level 3 is elusive, but there's enough software between purpose-built quality management applications, laboratory systems, MES applications and the automation and control area, not to mention ERP, to solve issues at all three levels. Are there other ways in which technology can support the goal of quality improvement, beyond deploying software solutions?

Jacobson: Software is a big component, but we must also look at technology in terms of collaborative capability. What I mean by that is the ability to create visibility or create some form of notification capability, not just inside the four walls of an organization but across multiple tiers of the supply chain.

This means being able to create portals or use instant messaging to help customers and suppliers collaborate in areas like process design, or even share knowledge on certain products and processes themselves. It also means extending out to customers -- the end user or the consumer -- to provide feedback on product use. So now we actually start to track quality across the lifecycle of a product, not within a silo of the supply chain. How can IT managers use technology to reduce or eliminate error rates?

Jacobson: I'm reminded of the quote I heard at a quality conference a couple of years ago from one quality manager, who said, "If we all did our jobs, we wouldn't be here."

The best thing for IT to do is provide appropriate data capture and propagation for any manufacturer -- whether it's someone at the line level, the plant operations level or the business level -- to act on the right event at the right time. And this means using IT to create situational awareness.

The best thing for IT to do is provide appropriate data capture and propagation for any manufacturer to act on the right event at the right time.

A fighter pilot may be the best example of this. A fighter pilot sits in the cockpit of a multi-million-dollar asset, and everything he needs to make split second decisions and tradeoffs in how to do his job is served up to him in a contextualized interface right in front of him.

We have to think about adapting that type of contextualized interface and that type of experience to the shop floor, to the management level and the business level. This way we can make tradeoffs between what we need to do at not just the site level from a production scheduling perspective, but also at the business level to understand how quality issues affect overall supply chain performance.

Case in point: If I have a failed asset tag that affects a batch, it means one thing to the maintenance worker, another to the production planner, and yet another to the supply chain. In other words, why the failed batch will prevent the execution of a profitable Perfect Order for the supply chain. So what's the logical starting point for manufacturing IT managers? Are there any special deployment issues?

Jacobson:The logical starting point when you want to look at how to apply IT to any manufacturing domain is to think about one thing: What is the business problem we're trying to solve? Too often, we go to whack in systems, to attack things such as variability, to attack Right First Time manufacture, to attack reducing scrap, rework and paper. And those are all great benefits.

But for the three levels of problems that I outlined, they work well only at level 1 or level 2. So when you start thinking about which business goals are going to drive these data and architecture efforts -- whether it's visibility to the supply chain, cost, or compliance -- you also need to start thinking about the level at which you're going to solve these problems. Then it becomes very easy to figure out which specific technological services you need to apply versus simply trying to attack with a systems approach first -- thinking about which systems lie in which little boxes. Isn't the systems approach the approach most manufacturers take?

Jacobson: We often think about that first because it's very easy to say we need to throw MES at this, but the reality is that we need to take a step back and think about which problem we want to solve. We need to figure out where our biggest bleeding point is with respect to quality. Then we can think about the systems approach that is needed to wrap IT around the issue.

In cases where you've already deployed IT, you may start to think about the issue of why IT isn't working there.

Next Steps

Analytics helps supply chain partners


This was last published in November 2009

Dig Deeper on Manufacturing production planning

Start the conversation

Send me notifications when other members comment.

Please create a username to comment.