A few years ago almost everyone you spoke to had a “bad China quality story”. Much of it was exaggerated and most of it was unfair given the huge number of high quality products “Made in China” including icons like the Apple iPhone, Airbus A320 and Mercedes E-Class. However the reality is quality problems do still occur with Chinese manufacturers.
I have written before about how western buyers are often the main source of their own “China quality” problems. This comes when buyers forget two critical points:
- You get what you pay for – by choosing unsuitable very low cost suppliers who do not have the capabilities or equipment to deliver a quality product.
- You get what you ask for – by failing to adequately document and specify products, failure to communicate fully and allow for language and cultural differences.
However beyond these issues there are some fundamental problems that I see in most Chinese factories that contribute to less than world-class quality performance. These problems can be seen in the piles of products being visually inspected in almost every factory.
Like western manufacturers 50 years ago, Chinese manufacturers have responded to their quality problems by putting in inspection steps. This has often been done at the direction of western suppliers who have demanded goods be inspected prior to shipping and often imposed onerous quality control plans incorporating detailed inspection.
Inspection does not Improve Quality
Unfortunately, as we were taught by W. Edwards Deming et. al. 30 years ago, inspection does not improve quality. Inspection just finds defects (or hopefully most of them). Humans are also not infallible and therefore if there are defects present then it is certain that some defects will not be seen by inspectors and will be sent to customers.
Worse than this, the presence of 100% inspection means that upstream operations learn that even if they make a few defects the inspectors will find them and stop them getting to the customer. As a result machine set-ups may not be done accurately, poor quality materials may be used and short cuts taken. In other words inspection can actually increase the number of defective products made (even though most defects will be removed by the inspectors). In one example in my Operations career in Australia we removed 100% inspection from a line manufacturing PET bottles and found that the quality delivered to the customer actually went up because defects were being prevented rather than relying on inspectors.
All this Inspection costs a Fortune
Costs in China are rising fast, driven by rising wage costs. Despite this I see many factories where 25% or more of their total factory labour are engaged in product inspection. This is a massive and avoidable manufacturing cost. Worse than that, if you give someone the role of inspector they will throw things out. Imagine if an inspector worked all day and did not claim to find one defect. I am sure his or her boss would be really wondering if they were doing their job. Therefore inspection guarantees a level of waste as products are rejected and either reworked or scrapped, in many cases unnecessarily. Add to that my previous point that end of line inspection creates a culture where “near enough is good enough”, increasing the number of defects created upstream, then you can see the cost of this inspection is very great indeed.
So What Should We Do?
It is not as simple as simple removing the inspection. Doing this will save some short term labour cost, but will immediately lead to the defects that the inspectors were finding going straight to the customer. Therefore reducing reliance on inspection is (unfortunately) a slow and painstaking process. There are three key steps to make this transition:
- Every Problem is an Opportunity: We don’t like problems. We usually like to find a quick fix and forget the problem. In quality this can mean simply scrapping or reworking the defective part. Instead we need to create a culture where defects are seen as opportunities to improve. This is tough in China where admitting problems is often associated with loss of face. However with the right language and approach it is possible to get an acceptance that “problems are good”. Every defect then becomes and opportunity to improve. For every defect we should be asking why it occurred, where in the process it occurred and what is the root cause. The TXM Solving Problems Every Day approach to structured problem solving using magnetic concern strips is a very effective way to achieve this. You may find one or two big causes of defects that can be addressed quickly, but more likely improvement is likely to come from dozens or even hundreds of little improvements addressed over months will add up to a major reduction of defects. At a joinery factory in Australia we used this technique to reduce defects by over 80% over the course of a year.
- Achieve Quality at Every Step: “Quality at the Source” is a fundamental pillar of Lean and the Toyota production system. This means that every person at every step in the process must have the tools, the capability, the skills and the responsibility to achieve perfect quality every time. Where they cannot achieve perfect quality every employee should be empowered to stop the process rather than passing on defective product to the next step. Attention needs to be paid to developing standard work so there is standard way to do every task that will deliver a quality outcome. Tools can provide tools to “error-proof” the task to prevent operators making mistakes such as mis-assembly or leaving out parts or ingredients. Upstream supplier quality also needs to be assured in the same way, not by inspection, but by choosing capable suppliers and working with them to eliminate the sources of defects in their processes. Products also need to be designed in a way to improve ease of manufacture and reduce the risk of defects being made in the first place.
- Capable Processes do not Produce Defects: Beyond the steps below is the concept of process capability. This is often associated with Six Sigma, but goes back much further than that. Essentially what this says that in a stable production process key quality variables will be normally distributed around a mean. In a normal distribution almost all possible outcomes will fall within a range six standard deviations either side of the mean (the Six Sigma range). Therefore if the six sigma range falls within the specification range the chance of producing a defect is almost zero (less than one in a million). As a result, provided such a process remains in control (no external factors cause large changes in outcomes) then defects will not occur and inspection will not be necessarily. I am explaining a complex concept in the simplest possible terms here, but if your normal range of outcomes falls outside your specification range, then you will make defects even if your process is stable. In this case you will need more sophisticated problem solving and process improvement tools (typically those used by a Six Sigma Black Belt) to improve process capability and therefore enable the inspection to be removed. In our experience though most quality problems can be overcome through the first two steps above.
Reducing reliance on inspection and building quality in to a product at every step of the process is a slow and difficult task. Ask most western Factory Managers and they will tell you as the west (lead by Japan) has been through this transition in the last 30 years since the birth of the “quality movement in the 1980’s. However achieving this change is essential if Chinese suppliers are truly to shake the tag of “poor quality”. My prediction is that this change will happen and it will happen much faster than it occurred in the west (in fact the mindset change is already well underway in advanced sectors such as automotive and electronics). TXM is already assisting companies to help their factories and their suppliers in China to change.