Q: I just accepted my first quality control manager position, and I seem to be running into a few problems. The biggest issue is that my co-managers are set in their old ways of penny pinching. I am trying to implement new and more productive methods and procedures, and I’m getting red-flagged by my one and only superior.
I have so many ideas for improvement but have no way of making others see my point of view. Slowly, I try to convince them of the pros of implementing these ideas. How do you negotiate with people who are afraid of improvement?
A: The best way to get someone to change is to answer this question for them: What is in it for me? To the extent that one can answer this question, getting them to change is that much easier.
You may want to try to translate "new and more productive methods and procedures" into financial terms. In other words, how much money will implementing new and more productive methods and procedures save? How much additional money will they generate, and in what time frame?
After being shown this in a clear manner, any reasonable person will want to make these changes. Try to explain pros and cons in terms of what benefits the persons being asked to change will get out of the proposed changes.
Some time ago, I conducted a quality-systems audit at a large, upscale, children’s-wear manufacturer in the United States. In the beginning, the director of quality mentioned to me he had a difficult time getting the necessary resources to manage quality effectively in this multiplant company. After a careful study, however, he showed senior management how much money was being lost due to poor quality and what effective quality management could save. Soon after, he got all the money and cooperation he needed.
Principal, Mehta Consulting LLC
Q: Are customer satisfaction surveys the only way to identify customer requirements and perceptions?
Dubai, United Arab Emirates
A: A formal customer survey, if done properly, can be useful in identifying customer requirements and perceptions. Its strong point is, theoretically at least, it can gather unbiased information, whereas relying on customer-initiated contacts will often generate a disproportionate number of upset or extremely happy customers.
Survey cards from restaurants and hotels also may get responses weighted toward the extremes. For example, I rarely fill out survey cards when I am generally satisfied with things, but if something really bothers me, I will be more likely to turn in a card. Yet survey cards are still useful, because they call attention to problems that otherwise could be overlooked. If the same problem continues to show up on the cards, it should be addressed immediately.
Many organizations monitor customer requirements and perceptions through product returns or complaints. It is important to make full use of these sources of information, regardless of whether formal surveys are also done. Some regulations require that all complaints generate corrective-action investigations unless a valid reason for not doing so is documented. Complaint data, coupled with surveys, can provide especially strong insights.
The usefulness of the survey is somewhat dependent on the product or service being considered. If a company makes large numbers of the same item or delivers the same service over and over, the survey can be a good source of data. If the company makes custom devices or very expensive products, however, it will want to nail down the customer requirements well before the product or service is delivered. A post-delivery survey may still be helpful, but it should uncover few surprises, if any.
Requirements need to be defined through specifications, drawings or contracts that are approved by the customer. Close contact between the provider and the customer throughout the design and delivery phases is far more important than the post-delivery survey. A post-delivery survey can still be useful in gaining insights into customer perceptions but should uncover very few, if any, surprises.
Even for a high-volume consumer product, where the customer may not be involved in the design, you should not rely entirely on surveys. A company should consider other tools, such as focus groups, analysis of data from previous designs and studies of competitors’ offerings.
Surveys can be useful in understanding customer requirements and perceptions but should only be a part of a customer data collection strategy.
Consultant, quality improvement and statistical methods
Fort Collins, CO
Q: Does rule No. 1 (of ASME Y14.5M-1994) apply to injection molded plastic parts given the free-state exemption?
A: Rule No. 1, in one sentence, says "when only a tolerance of size is specified, that tolerance controls both size and form." You seem to be concerned with the form end of things.
The first question is whether it applies to injection molded parts. Yes, rule No. 1 applies to injection molded plastic parts, but it applies to all parts, whether they’re injection molded or not. You say, "given the free-state exemption." Here’s the bottom line. The exemption deals with two things: rigid parts and nonrigid parts. Whether it’s a rigid part or a nonrigid part has nothing to do with injection molding. You can have a rigid part in injection molding just as easily as a nonrigid part. The exemption applies to nonrigid parts.
The first thing you do is go back to the print. The drawing should specify the drawing standard the print is drawn to. If it’s a geographic dimension and tolerancing (GD&T) print, then it should be ANSI Y14.5M-1982, reaffirmed in 1988 (what I call the old standard) or the revision to it, ASME Y14.5M-1994, reaffirmed in 1999. Those two standards, depending on what the age of the print is, will help you determine which way to go.
I don’t see huge differences between the 1982 standard and the 1994 standard. In many areas, they don’t conflict with each other. If you’re dealing with something like symmetry, you need to know which standard you’re working with. But with a situation like this, I don’t think it really matters.
The whole key is whether the part is rigid or nonrigid. If you’re doing injection molded parts and the parts are rigid, then you need to check them in the free state. If the part is injection molded and nonrigid, it’s specified in the print what the tolerance is for the amount of free-state variation allowed.
The design engineer can put a callout on the print by GD&T that has a circle with an "F" in it. This means you’re allowed free-state variation. If there’s no "F" on the print, you need to check it within the free state, and free-state variation isn’t allowed. Here’s the rub: Design engineers aren’t going to waste their time putting a free-state variation on something like O-rings, because everyone knows they’re a nonrigid part.
Because the design engineers sometimes just assume you know what you’re doing, if it’s a flexible part, you would be allowed to take the part and move it to the specified data of the part, clamp the part in place according to those data and then check it. Basically, what you’re doing is taking the free-state variation out of the part, moving it to what its rigid envelope would be and checking it in place.
In some cases, it may be required for the part to meet its tolerance requirements while in the free state. In other situations, it may be necessary to simulate the actual-use conditions of the part, warp it back into where it should be, hold it in place and then check it. This is done by restraining the appropriate features, such as the data-reference features.
The restraining forces are those that would be exerted in the assembly or in the function of the part. Let’s say it takes 400 pounds of pressure to put a part back into shape and check it. The spot weld isn’t going to hold 400 pounds. But if I take four pounds of pressure to bring it back into shape, a spot weld can hold that. Nobody’s going to care about me doing that. You can’t use excessive force. That’s why, in ASME Y14.5M:1994, it says the acceptable forces are those that would be exerted in the assembly or functioning of the part.
Gregory S. Gay