Quality: So what good is it anyway, from an operations perspective?

Quality: So what good is it anyway, from an operations perspective?
August 6, 2013 Frederick S. Buchman

3d-financial-graph-2-1002066-mOkay, we know that ‘Operations types’ focus on production, yield, shipping/receiving of raw material, scheduling, and more production.  They figure their job is to make product, and more product, and continue making product according to the Master Schedule (at the required yield level) in order to make their numbers for the day/week/month.  And they are right.  No doubt, we need these people to do their very best to meet demand, produce to the schedule, and deliver everything.

So where does ‘quality’ fit in?  That is the focus of this article:  Deriving a business advantage of applying a quality approach, such as how quality can enhance and augment the ability of Operations to meet and even exceed its production goals, while also reducing (yes, reducing, not increasing!) costs and satisfying customers even more than just by standard performance.  We will address Yield first, then ‘batch’ and ‘work-in-process’ (commonly referred to as WIP) inventory. Lastly, we will discuss failures in first-pass-yield as we talk about how integrating quality with production can indeed enhance operational performance and results.

1.  Let’s start by looking at ‘Yield’, the almighty KPI of operations.  Clearly, with perfect quality, we make exactly what we ship.  A 5% quality yield failure requires we make 5% more, in order to meet demand.  As yield failure increases, so does the required production volume.  Yield failure can occur from defective material, sub-components or assemblies, improper process steps, or handling damage.  We react to each of these by increasing the number of units produced, and actually take these failure rates into account when we design our manufacturing processes.

But would it not be better to avoid adding volume to compensate for this problem?  Using quality tools, we can study the chronic behavior of these failures to see if they are predictable, and therefore, preventable.  In many cases, we even know they are happening and have built-in rework in anticipation of their failure, all of which adds cost and time to the delivery cycle.

Let’s take an example from the Semiconductor industry:  Chip yield suffers within each wafer due to a specific rate of failure in comprehensive circuit testing while still uncut.  Further yield casualties occur once the chips are cut out, and each applied onto the copper substrate and ‘wire-bonded’.  Additional test failures necessitate manufacturing even more chips, with the added task of melting down failed units for reclamation of the gold and other materials for reuse.  All of this imperfection and failure in first pass and subsequent yield requires 50% or more over production to compensate and still meet delivery requirements.

The application of Lean, Six Sigma and other CI tools can provide the insight relating to the behavior of both the design and production process that is contributing to these failures, whether chronic or special in nature, to enable reduction and/or elimination.  Trend information can reveal error causing production tendencies, which can be corrected to increase yield.  Cp and Cpk shows the capability of a process or product in order to help understand delivery limitations.  Other tools focused on the flow of a product or service can show where waste in inventory, bottlenecks and impediments exist to restrict process efficiency and yield.  All of these tools and techniques can help to increase high quality production while even reducing costs associated with rework, reclamation of precious materials, and other wasted, non-value related activities that many companies currently engage in to meet demand.

2.  Now let’s look at the manufacturing process itself.  From an operational perspective, manufacturing processes are designed to produce high quantities of a product or component.  One ‘core’ belief is that we need to batch as much as possible at each step (concentrating on high rate production) to minimize travel, and thus reduce the per-unit cost.  As a result, we design-in very high in-process capital costs based on ‘work in-process’ inventory.

For example, a wrench company I worked with some time ago would fill large drums with several thousand wrenches to be worked on, and subsequently moved to each manufacturing process step.  Of course, the actual number of wrenches being worked on at each station was only a fraction of the batch, so the entire barrel of wrenches had to wait a while before it moved to the next step.  As a result, while the actual work to manufacture and finish a wrench took maybe a couple hours, it would take the batches several weeks to be completed.  With only one type of wrench, and if everything stayed exactly the same, this would be fine.  However, as we know, things don’t stay the same, or even consistent for that matter.  For the wrench company, they started to see significant problems when the customer started changing their orders, ordering different models in different quantities.  As more models come into the picture, the greater the need for flexibility becomes; and the worse the problem would get since the batch process requires you to make an entire batch of each type of new product. This alone vastly increased both the work-in-process and finished-goods inventory, and necessitated building an additional warehouse in order to house all of the excess finished wrenches.

It came to a crisis, however, when a critical material problem came up.  The material issue was not discovered until some type of end-inspection (a typical procedure with mass production lines), at which point vast numbers of wrenches had to be melted down and remade, not just from the production line, but from the finished goods inventory stores as well.  This created huge losses for the company, and almost closed it altogether.

Given the ‘batch’ mentality, one can see how easily it is to get buried in ‘work in-process’ inventory.  Add the potential for material failure, and you suddenly have a lot of product you can’t sell or have to melt or convert back to raw material, thus losing any added value and residual profit.  Many companies today are currently wrestling with this problem, and have to charge more to cover the cost of this unnecessary work in-process or non-conforming inventory.

A related issue with traditional manufacturing methods relates to organizing a production group by function.  For example, in a traditional connector plant, all stamping, plating, molding, and other such functions are located together (possibly even in separate buildings) with the product ‘queuing’ to wait for the next available machine as it is produced and packaged.  This is another form of ‘batching’, but by process, not product or component type.  The chemical bath and tooling set-ups are time consuming, and are specific to each connector type and model.  We can begin to see the weakness here as well – with customer order changes in plating type, molding material and hardness, and other possibilities, it becomes very difficult to keep up with demand given the set-up requirements and the difficulties with changeovers where chemicals are involved.  And again, should there be any material issues, they would not be discovered until the connecters were in final assembly and tested, and then massive quantities would have to be disassembled, melted down and reused – all waste that is a direct result of the design of the manufacturing process.

The need for greater manufacturing flexibility points to a need to change the way things are made, not by batch, but by small numbers in many manufacturing cells, each producing in parallel in order to be able to flex to demand changes.  Moving to this lean model would enable products to be manufactured, finished and shipped in a fraction of the time of batched process cycle times.  This would also greatly reduce the amount of work-in-process inventory, and thus capital costs to produce (i.e. per unit cost).  But there are additional benefits as well.

If your customer needs your product regularly, say every other day, but your process produces it in 2 weeks, this customer would have to purchase 2 weeks-worth of inventory in order to be able to support their needs.  But if you could reduce your delivery cycle time by producing your product in 1-2 days, then your customer could receive only as much as they would need with a 1 day cushion. This would drop their inventory costs by over 80% and reduce storage space requirements, not to mention handle damage, loss, and unnecessary added movement and tracking costs which ultimately enhance your customer’s satisfaction with your product and delivery performance.  And if there are any material or component issues, they are quickly replaced, with minimal loss.  Also, if your customer suddenly changes requirements, your plant can react in real-time, instead of trying to figure out how to dispose of 2 weeks of inventory.  These are just some of the advantages that can be gained through quality and lean methods as applied to the production process itself.

3.  We want to understand not just the ‘yield’, but what we call the ‘first pass yield’, at each station, in each process step. This ‘first pass yield’ (often referred to as ‘rolled throughput yield’) is the rate of incidence where the process step or station activity was performed perfectly, as planned, without any need to adjust, react, or correct.  While we often see a high level of ‘yield’, the ‘first pass yield’ rate is very often less than 1/10th or about 5-8% of the yield.  Which means that about 93-95% of the work done requires some adjustment or correction to be accomplished successfully.  We sometimes call these adjustments ‘work-arounds’, where we fix, adjust, clean, rearrange, pre-treat, set up, and correct incoming material or components in order to be able to complete a specific activity in the process.  Many times, we design these added preparatory tasks into the manufacturing process, or add them depending on in-coming material/component condition.  The key here is to assess each input to the next step, and determine if there are any of these ‘preparatory’ types of activities that are being done prior to the actual value added work, if any, and see if there are other ways to increase the first pass yield at that step in the process.

Of course, it would be desirable not to have to do any of these things at all, but often they become part of the process, no longer visible as waste generators, but rather just another activity required to produce and deliver your products.  It is these ‘work-arounds’ that we target with specific quality and lean techniques where we make them ‘visible’ again and see them for the waste they create, not to mention the added cycle time, cost and rework.  The key is to assess each task to see if it adds value or if it is just something we do when something can’t be accomplished right the first time.

The definition of ‘adding value’ is very specific in this context, and can only be attributed to those tasks that actually change the product or service, and that would be valued and even paid for in the eyes of the customer.  Given these criterion, it becomes clear that there are a multitude of activities involving receiving, inspecting, transferring, storing, tracking, reworking, overproducing, and other things that are important to the process, but not to the customer.  Many of these activities are designed in, or are the result of a need to compensate for inadequate incoming material condition, such as: burrs on edges that have to be removed, forms with headers that are not completely filled out, or surfaces that have to be pre-cleaned prior to treatment.  We ‘build in’ mini-rework stations to prepare parts, components or materials prior to use rather than ensure we get them at each step exactly as usable.  And because they become ‘routine’, we eventually stop asking, but rather just see it as another step.

Once we have identified the added wasteful activities, we can attack the root cause of why we are doing them, and eliminate the need, or at least drastically reduce it.

As you can imagine, eliminating or reducing these wasteful activities would dramatically improve the ‘first pass yield’; reduce the cycle time of the process and thus operational costs; and improve quality and overall yield.  In addition, this would also dramatically reduce ‘work in-process’ inventory levels, and enable much better station performance results.

These are just three of the many ways that quality tools and techniques can help operations achieve and often exceed its production objectives.  You have only to look at the process, and the way you manufacture and deliver your products to see where the opportunities exist, and then you can begin to attack those root causes that rob you of your true performance entitlement.

So the next time someone says:  “Quality – what can they do for us?”, you have a real response, one that makes sense, relates to what is important to operations, and can definitely deliver in improved performance and results.

By Frederick S. Buchman


Working in Operational Excellence over 25 years, Frederick S. Buchman is the President and CEO of Hayward Enterprises, Inc., and is a recognized international author, speaker, coach and consultant on Operational and Process Excellence programs, co-authoring multi-language books on systems for management such as balanced scorecards and dashboards.  His organization has helped many Fortune 100 companies worldwide in successfully designing, executing and integrating their continuous improvement programs, and also often assists with senior leader qualification, executive team development, and working with companies big and small to achieve and sustain their performance goals. 

Listen to our free monthly podcast!

Listen
SUBSCRIBE TO OUR NEWSLETTER