What Does Quality Mean to Chinese Manufacturers?

I often ask potential hires “what is quality?” and it seems even people who’ve been in the quality assurance industry for 10 years don’t have a ready answer.

It seems like I am not the only one. Here are the most common responses gathered by the American Society for Quality during a recent survey:

quality definitions

I guess one could argue that there are several definitions, depending how high up in the sky we choose to be:

  • At the micro level: “meeting the customer’s specifications”
  • In a slightly less narrow view: “fitness for use” combined with “never stopping to improve processes”
  • In terms an accountant understands: “ensuring customers come back and products do not”
  • On a more macro level: “doing the right things in the right way through the practice of excellence”

Unfortunately, most Chinese companies are still at the micro level. Producing a quality product means complying with rules imposed by a customer. With that mindset, higher quality means extra inspectors, who cost more money.

I wrote about this in Is it expensive to increase the quality level in a factory?:

Importers use statistical quality control standards based on AQL limits. If a buyer sets a tolerance tighter than what is usually considered “normal” for general consumer goods, the supplier generally raises prices.

What is wrong with this mindset? It is 100% reactive, 0% proactive. There is no initiative to improve design & production processes.

Maybe some readers can propose a definition of quality that everyone can agree on?

Mobile Applications (on Tablets) for Quality Inspections

Over the past three years, I guess more than a dozen people have told me “tablets are a perfect fit for inspectors”. Yet I haven’t seen any company that developed a good application that works on iPad or Android, and that is suitable for quality inspections.

The truth is, it is complicated. We’ve been working for 4 months on an app and it is only starting to work nicely. And that’s without counting all the work on the IT system to be used in the office — that was another 8 months of development…

And yet, this is way to go for all inspectors. There is nothing that is currently done on paper or on a laptop that can’t be done on a tablet. But there are LOTS of things that can be done on a tablet and that are currently impossible for most inspectors.

A good tablet application brings a lot of benefits to quality control managers:
  • Statistics about suppliers and inspectorsŠ
  • History of inspector actions with exact time, to ensure they take the time necessary to do the job and to understand what sequence they follow
  • Assurance that right files will be available
  • Real-time information for urgent orders (contingent on internet connection inside the factory)
There are also benefits for the quality department¹s admin assistant:
  • Easier management of samples, tools, etc.
  • Many semi-automated actions (confirmation emails, requests for tools, etc.)
And, last but not least, the benefits for QC inspectors are tremendous:
  • Reporting, sending photos, etc. takes much less time (it is all automated)
  • They are guided step by step in a procedure (lower risk of forgetting something)
  • Claiming expenses takes less time
  • A tablet is lighter and cheaper than a laptop

So here you go… This is the future. And maybe Google Glass will be even more convenient. I’d be curious to head from readers who have already experimented with mobile applications.

(By the way, any self-promotional comments will be erased.)

Moving Tooling/Molds from one Chinese Factory to Another

Clients asked us several times to oversee a transfer of tools/molds from one factory to another. So we have written a checklist (see below).

Why move tooling from one factory to another?

  • Maybe the buyer selects one factory to make a mold and another one to do the injection. This is generally not advised, but sometimes the buyer has no choice.
  • Maybe the business stops with the original factory, which accepts to release the tooling (it doesn’t always happen, and a good contract certainly helps).

In this article we are making the assumption that the tooling is a mold for plastic injection, and that this mold needs to be moved from the toolmaker to the molding factory.

For those importers who want to minimize risks, the sequence to follow is as below:

1. BUYER: – All tooling should be signed off as complete and ready for use in production. Documentation should be in place showing the buyer has signed off each tool.

2. BUYER: – Tooling transfer plan including schedule should be generated and in place with all parties (the buyer, the toolmaker, and the receiving factory that will run the tools in mass production).

3. TOOLMAKER: – Tooling transfer documentation should include all tooling 2D drawings as well as 3D CAD data, release and transfer of tooling contract, preliminary settings for each tool as a guide.

4. BUYER/INSPECTION: – Buyer’s engineer or third party representative should check each tool against the transfer contract to ensure everything is correct and accountable for.

5. BUYER/TOOLMAKER: – Transfer readiness should include rust protection, general packaging protection, the correct shipping packaging (depending on shipping method, land, sea, or air). All this needs to be documented and checked during the packaging stage.

6. BUYER/TOOLMAKER: – Transfer of tools, depending on distance and method of transport. If local, the toolmaker may use their own truck, in which case the buyer’s engineer or third party representative should accompany the tools during transportation. If longer distances, the buyer is advised to arrange shipment through their own freight forwarder.

7. BUYER: – Ensure all documentation has been sent to the receiving factory ready for acceptance of transferred tools.

8. MOLDER/BUYER: – Upon receiving transferred tools, unpack tools and check off inventory to ensure everything has been delivered.

9. BUYER/INSPECTION: – Buyer’s engineer or third party representative should check each tool against the transfer contract to ensure everything is correct and accountable for.

10. MOLDER/BUYER: – Buyer’s engineer or third party representative should work with mold factory and run each tool in order to obtain samples for inspection.

11. BUYER: – Buyer’s engineer or third party representative should inspect initial sample and cross check against signed off samples from the toolmaker.

12. BUYER/MOLDER: – Once all tools have been accounted for, checked they are in good condition and are able to run in the factory’s machines, the buyer and the molding factory need to sign tool transfer contract. Through that contract they accept all tools, take responsibility for them, and acknowledge the buyer’s ownership rights (among other clauses).

13. BUYER: – Buyer’s engineer or third party representative should check storage facilities to ensure each tool will be stored correctly and safely and is easily retrievable when needed.

14. BUYER: – Buyer’s engineer or third party representative should check tool maintenance capabilities to ensure adequate skills and equipment are in-place to maintain the tools at the highest quality.

Now, what if the tools are moved from an old (and bad) supplier to a new supplier? The old supplier probably won’t be as cooperative, so some steps will have to be skipped. And the buyer is advised to arrange transportation through his freight forwarder, to avoid contact between the 2 suppliers.

Maybe some readers can offer a few more tips for this type of situation?

How to Use Statistical Tools to Improve Production Processes

In Limits of Statistical Process Control in China, experienced consultant Brad Pritts described his observations over the years. Below is his advice to use statistical tools to improve production processes.

 —–

What I do when I work with companies, whether US or Chinese, is the following approach.  I learned most of this from books by Dr. Don Wheeler, a statistician and disciple of Deming.  I strongly recommend anything written by Wheeler to people who want to use statistics to improve quality.  While I am a skeptic about Xbar/R, I am an absolute believer in the value of statistics for process improvement.   Wheeler’s books do a good job of explaining variation without unnecessary complexity.

First, determine the capability of your measurement system.  This is done using gage repeatability and reproducibility studies, as well as determining bias and linearity.  No measurement system is perfect.  You need to know that your measurement system is effective relative to the degree of variation your process has, and the tolerances you need to achieve.

One mentor I worked with emphasized the importance of this point by stating emphatically that if you changed a manufacturing process based on unproven measurements, you were guilty of engineering malpractice– tinkering
blindly.

Many times it’s necessary to spend quite a bit of work improving the measurement process before tackling the production process itself.

Second, determine whether your process is stable in the short run.  Take a sample of, say, 20 or 30 consecutive pieces.  Chart the characteristics as a simple run chart. Look for unnatural patterns in the data – trends up or down, sudden shifts, etc.  Calculate the short term process capability.

Often this step alone will show you where process improvement is needed. Sometimes it will be very simple things — tightening up loose fittings in a machine, making an adjustment to a fixture.  Other times it may be very difficult – for example, it may be necessary to redesign and rebuild a stamping die… or replace a machine with a better technology machine.  These may happen but only with major management commitment.

Third, use the knowledge of the process to establish inspection frequencies in the control plan. Highly capable processes may only need to be checked at first piece setup while troublesome processes may need 100% inspection.

Finally, where warranted, do ongoing longer term studies to see what is changing over the long haul — due to machine and tool wear, changes in operators, different batches of raw material, etc.

One amusing project I was involved with was an assembly operation. Simple statistical analysis helped us in an unusual way. This product had a 32 person assembly workforce.  Most of the workers had been newly recruited, as this operation was a new job to this company. Turnover was high.

We charted the defect and rework rates and found a striking pattern in defects which were clustered on occasional Mondays.  With a bit of discussion we found that the defects resulted from a practice of hiring several new workers at a time, and starting them all on the line on Monday mornings, overtaxing the supervisors’ abilities to train and monitor their work.  We changed the hiring practices to bring new people in on a staggered basis.  No new people on Monday morning!  Tuesday, bring in one new worker on a half-day shift (starting mid-day), then progressing to a full shift on Wednesday;  if multiple workers were needed a second new worker could be added Thursday in the same pattern.  Defects and customer complaints fell by about two thirds!

While I am proud of my education and experience with complex statistics, the fact is that many of my biggest success stories come from very simple, practical moves like the story above.

Conclusions:

  1. Reducing variability is a key to improving quality and profit.
  2. To understand and reduce variability, statistics are a powerful tool.
  3. Use the simplest statistics first — counts, totals, average measurements, run charts. Sometimes these are all that are needed, and you’ll have a much easier time explaining your results to others.
  4. Confirm (and if necessary, improve) your measurement processes first. Only then will you be able to effectively measure and improve the actual production process.
  5. Assess short term and long term process capability, and use this knowledge to set up your inspection program and improvement priorities.
  6. If ongoing Shewhart style x-bar and r charts are needed, either automate them, or plan on extensive supervisory and worker training, handholding, and management if you expect them to work.   

Particularly critical is establishing an environment where workers are willing to honestly report problems. This is difficult in many companies, American or Chinese.  If you can’t get to this situation don’t waste your time on SPC.

Limits of Statistical Process Control in China

I am always interested in tools that prove useful in improving a factory’s operations. But I always wondered, how useful is Statistical Process Control (SPC)? The fact is, I have never seen a Chinese factory make good use of SPC tools.

I asked Brad Pritts, an experienced quality consultant specializing in the auto industry, for his opinion. Some classic SPC tools are listed in the “core quality tools” of the North American car industry. So it would make sense to have lots of factories apply SPC… Or so I thought.

——–

Here is what Brad wrote back:

In thirty-five years of quality consulting I have worked with or audited about 150 companies. Most of these were medium to small size auto parts manufacturers (25-250 employees), and most are in the USA.  However, I do include about 20 Chinese factories in the list.

Auto parts manufacturers have been mandated by their customers to implement SPC since the late 1980’s;  so most of these companies made some effort to use SPC. My comments below are based on my experience in this particular environment and are not intended to condemn SPC in general.

Many companies have made good use of statistics to manage and improve quality.

Very few in my limited experience have made effective use of textbook X bar/R or X bar/ moving R charts (Shewhart charts) to manage “online” process control. I can think of only two companies in this category.

  • One was a steel processor who monitored several characteristics on their pickling and slitting operation, as coil steel flowed through a machine at fairly high speed.
  • The second is a fastener maker who has an end of line 100% automated inspection of bolts that uses ongoing inspection data to identify and automatically reject (“kick out”) individual pieces that fall outside control limits.

Neither of these companies was in China.

Both successful cases had some common threads.

  1. Both used the SPC in automated, or nearly fully automated, implementations. The measurements were taken with automated instruments;  fed directly to a process control computer;  and the calculations fully done by machine.
  2. Both had relatively large, uniform volumes of material flowing through these processes. (This helped cost justify the large investment in technology.)
  3. In both companies, the systems were implemented by a few really clever engineers, and were accepted but not really understood by most of the employees. (In the fastener maker, I am curious to see how, or whether, the system will survive when its designer retires next year!)

Besides these cases in my personal experience, I can relate a few cases described by a colleague who sold and installed process control equipment.  He had a number of successes with automated process monitoring of large scale industrial processes such as paper-making and steel rolling mills, where the control equipment was programmed to use SPC to directly control process variables.

They were able to show process and profit improvements by first improving process capability (often reducing scrap and waste) and then optimizing the process.  You can picture these cases as having the same common factors – highly automated, large volume processes.  The capital investments for the process control systems were typically in the $100,000 USD range or more.

On the other hand I have seen a lot of SPC (traditional Shewhart Xbar/R) implementations which were a total waste of time and money, done to create a show or to meet a customer mandate.  Some of these were done with all good intent while others were frauds.

These had some common features, too.

In many cases employees gamed the system by not documenting out of control conditions, but instead stopping, adjusting the process, and taking new samples once the process was back in spec. (Note: in spec, not in statistical control.) A good analyst can usually detect this in the SPC data, because this behavior will prevent the system from attaining stability.  Meanwhile, the operators and management will conclude that SPC is worthless. (Which is true if you operate it this way.)

In some cases QA people gamed the system by selecting a few variables which weren’t really important to chart but in statistical control. so that they could report “good” Cpk values to customers, while not charting difficult but important variables. Another waste of time, although I could rationalize this as a sales and marketing expense — giving the customer the appearance of what they demanded.

One of my favorites was a company which installed a mostly automated system with gages wired directly to a computer monitoring system.  The sampling plan was set to 5 pieces per hour for X-bar/R charting. The operator was supposed to sample 5 consecutive pieces, put them in the gage and send the results to the computer which would do the charting and report a result. Data was transmitted by putting a part in the gage and pressing a foot pedal to send the data to the computer. To “save time” some operators would put one part in the gage and then press the foot pedal 5 times, pretending that 5 parts had been checked. As you can imagine when you do this the part to part variation is zero. But, the variation for the piece run the next hour will appear awful because it’s measured compared to the zero piece to piece variation! Again, a good SPC analyst can figure this out if they study the data (and surreptitiously watch the operator!)

One reason for  the difficulty in application is that many of the companies I have worked with have processes that are equipment or tooling dominant.

Think of metal stamping, plastic injection molding, or cold heading of fasteners. Ongoing realtime SPC helps little in these cases.  SPC may work much better in machining operations (detecting and controlling tool wear)
but I haven’t had much personal experience in that industry.

The next article will be Brad’s advice on how to use statistics in a useful way to get processes under control (click here).