Quality Control - Fishbone Diagrams

One of the simplest  tools to evaluate cause and effect in quality control is the fishbone diagram.  It is simple to understand and can be implemented on a piece of paper if that is all that you have.  From it you can often quickly determine a root cause of a problem and then take the necessary steps to correcting the problem to move forward.

Using the analogy of a "fishbone", the "head" of a fish is identified as the cause of the problem.  Working back toward the tail of the fish, potential causes of the problem are identified.  Each potential cause is carefully and thoroughly evaluated to verify if is truly a cause of the problem.  As the causes are evaluated, the diagram is filled out to look like skeletal structure of a fish.

(Fishbone Diagram template - courtesy of resourcesteaching.com)

I have used this type of analysis several times and it has been proven to be very successful.  There is no advanced training or education necessary to understand it.  Only a willingness to fix a problem.

Change Boards in the Operations Process

In the operations world, no design becomes static once it is released. Although most designs are well developed before initial release, there are a number of reasons why a design must be changed:

• The released engineering is difficult to manufacture, causing a modification or change to dimensions, materials, or process.
• The design is costly to produce and cost savings programs dictate that it needs to be produced at cheaper cost,  This also applies to weight savings (i.e aircraft) where extra weight hampers technical performance and/or fuel consumption.
• Other reasons that improve the manufacturability or delivery of a part, such as inspection criteria, inclusion of new reference notes, etc.

(Courtesy of mountainvision.blogspot.com)

To get operations and engineering to agree on a new go forward path, a change board is set up.  The change board can be used for any size company but is especially necessary in a larger company (greater than 1000 employees).

To start the process, a kickoff meeting is set up to explain the "problem" and outline the solution to it.  This is usually initiated by engineering or manufacturing because one or both of these two groups will be the catalyst to the change.  Invite representatives from any affected group to attend the meeting.  Inviting "extras" doesn't hurt: they will leave the meeting if the change doesn't affect them.

During the meeting explain the change in full detail, especially explaining the cost/ schedule benefits to proceeding with the change.  Upper management will want to see "the numbers" and will balk at changing the status quo if there is no a good case for it.  After all have given input (hopefully they all approve the change), then the schedulers can decide when the change will be implemented.  Of course good communication to the group is essential with PowerPoint slides and other documentation shared along the  way.

A successful change board meeting will please eveyone!
(Photo courtesy of chaicagocitylimits.com)

Engineers on the Shop Floor

(a Boeing 747 being assembled - courtesy of hotcharchipotch.wordpress.com)

In an earlier post I explained how it is essential to have the base of engineering operations as close to the manufacturing operations as possible.  In a ideal case, would be to have the engineering and manufacturing in the same building or in buildings next to each other.  Proximity to one another is what makes a successful operation.

But it is not enough for engineering and manufacturing to be near each other.  The organization must make an effort to encourage (hopefully not force) engineering to spend time on the factory floor.  There they can learn about how different stages of the operation work and can spend time talking to mechanics to learn what their challenges are in fabricating and building parts.  At Boeing (and at other large companies I would assume the same), young new hire engineers often do not want to visit the factory and learn.  Because of organization attitudes they are led to believe that engineers design the product and "throw it over the fence" to operations to build.  Not very efficient.

On of the best ways to get engineers to engage the factory is to put them in a liaison engineer role.  The liaison engineer is a degreed engineer who is works on the factory floor and helps bridge communication between engineering design and  manufacturing.  Often the liaison engineer has authority to repair parts damaged during manufacturing which expedites operations immensely.  At Boeing I was a liaison engineer for a little over a year and that experience made me a better design engineer by learning how to design a part that could more easily manufactured and later installed in the aircraft.

(Engineers on the factory floor)

I served as a liaison engineer later in my career and would advise organizations to schedule the new hire engineers to rotate as a liaison engineer after 1 - 2 years design experience.  An early exposure to shop practices will make them better design engineers which improves the entire operation.

Just-in-time

One of the best ideas to come out of operations management is the concept of just-in-time (JIT) production.  As most know it is the idea that whatever your operation is waiting for (raw materials, finished product, etc.), it will be there right when you need it.  This avoids keeping a large inventory of items on hand which must be kept track of.  With careful planning a just-in-time system will give you exactly what you want exactly when you need it.

(Courtesy www.planproduction.tripod.com)

Of course it doesn't always work that way.  It has been my experience that some products will invariably be delayed to the downstream user because of raw material shortages, errors in production, having other work "out prioritize" your work.  Instead of a just-in-time system, you know have a "just-in-turmoil" system.  With one delay in the pipeline the rest of the operation must now work overtime to catch up.  Literally work overtime as in a 24/7 type of rescue.

The best approach is to have the parts or materials arrive a little before "just-in-time" so there is a a little safety buffer.  It may mean stacking parts in side areas a few days before you need them. This may be uncomfortable, but will keep your operation flowing smoothly.

(Courtesy: www.elblogsalmon.com)

Rapid Prototyping

Rapid Prototyping serves a fantastic role in the product or service design stage within operations.  Historically a designer would do the best job possible in identifying areas of concern prior to releasing a design to be fabricated.  Once the design was released it usually went into fabrication and (with fingers crossed) everything went well.  If not there would be a revision to the design and the process started over and more finger crossing.

With the advent of rapid prototyping, it was easy and fast to create a first look at a design before it went into production.  All the designer has to to is send a 3D CAD model to the prototyping shop.  Then the programmers would write the program and have the part "built", often using stereolithography to produce it.  Then the part can be inspected by all concerned parties and any modifications can be done quickly.  This means it can then be implemented quickly in time to support a tight production schedule.

During my time at Boeing, I was fortunate to have visited the rapid prototyping shop in South Seattle.  They could take any complex design and fabricate it to reveal what the part would look like.  It was fast and the results were amazing.

These two chess pieces were made at a rapid prototyping shop.  The grey piece made from a steel alloy and the red piece is made from a plastic.  Notice the detail on the close up of the red piece below.

The popularity of 3D printing may eventually take the place of a prototype shop.  But until everone has a 3D printer, it is worth the time to use a rapid prototype shop to prevent design delays during a critical part of the operations flow.

Plant Audits

Arriving at a supplier's plant to audit the operations process is an activity that causes much anxiety for the supplier (at least the first couple of times).  Generally it is done when there are problems identified in the flow of operations.  There may be office related (i.e. engineering) or  production related (i.e. manufacturing).  Both areas must be addressed.

(Courtesy: www.pooltechservices.co.uk)

An engineering audit may uncover errors in design.  I have audited several supplier's while working at Boeing.  One supplier had a different type of CAD software than Boeing was using (that in itself was a problem).  This particular supplier had several issues with the 3D models that the Boeing team helped uncover that might have been missed if not for the audit.

All specifications and inspection criteria must clearly be flowed down from the prime contractor to the supplier.  I saw this issue happen during the same audit trip and this time it was Boeing's fault that the correct specifications were not imparted to the supplier.

It's also valuable to ensure that the supplier's engineering team is in sync with its manufacturing team. You can usually tell how this relationship is going during a kickoff meeting day one of your plant visit.  If relations between the two appear distant or aloof, then there is trouble.

For the manufacturing side, all aspects of the process are reviewed.  Do the machines break down often?  If the break, how long is the downtime?  Are the mechanics and machine operators properly trained?  What are the specific training plans for each operation?  Is the layout of the plant optimized?  All questions that need answers.  And that is just the start.

Some people are afraid to respectfully challenge the supplier.  During my audit trip, one of my fellow Boeing engineers continually challenged the supplier about all aspects of production.  He did it in a polite but assertive manner.  But the supplier eventually got tired of him and later told him he was not welcome back to the supplier's plant.  That engineer eventually became a Boeing manager because he embraced due diligence in seeking to make the operation more efficient.

Typical Supplier Manufacturing Plant - courtesy www.logicpol.com)

Real World Use of Statistical Process Control (SPC)

(Courtesy: wallpapers-xs.blogspot.com)

If you read many books on lean six sigma, you quickly realize that much of the focus is on data collection and analysis.  If you are like me you wonder if there are any real world applications and how can using the data collection techniques improve a process.

Early on in my earlier life as a design engineer at Boeing, I was part looking into a tolerance study of stringers on the B-2 bomber program.  These stringers were attached inside the wing and serve to transfer load from the wing skin to the underlying frames and ribs.  There were issues with the location of the stringers being out of position and our engineering group sought to determine why the stringers were out of position.

A generic configuration showing the long slender stringers as they sit in a wing box - courtesy of avcom.co,za 

We took location measurements of where the stringers were actually resting on the wings stored in the factory.  We then used the +/- tolerances as the upper control limits.  We then gathered data and plotted the stringer locations as measured to try to determine if there were any significant common trends (i.e. at a particular stinger number, at  certain coordinates in a wing, etc.)

Unfortunately I left that group shortly thereafter and I lost track of the final outcome.  At that time six sigma was not nearly as accepted as today, so I don't think we even knew what else to do and how to go into much greater analysis.  But I gained a quick indoctrination into six sigma and know it can show how to improve a process or product if used effectively.