APPENDIX E
Interview Summaries
During the Fall 2001 semester, the EE 491 group interviewed a variety of industry experts found among the ISU faculty. This was an attempt to gain some realistic knowledge about decision-making. This allowed the group to better understand how decision-making is related to every day operations.
In general, the interviewees spoke about making how they made decisions during their industry experiences. They had varying degrees of knowledge about formalized methods and algorithms. They all spoke about the ability to use more general methods to make decisions quickly and cheaply. This method was compared to the algorithm approach to show advantages and disadvantages.
Another common topic was teamwork and how people interacted during decisions. Several interviewers discussed the necessity of having the correct arrangement of people involved on a decision and the types of conflicts that often arise due to personal problems.
The following notes are provided from several of the interviews that were conducted.
Notes from interview with Doug Gemmill (8 November 2001)
Student: What has been you experience with the decision-making process?
Prof: Mainly teaching. I teach systems engineering. Which is sort of the process for developing a product. The decision-making that I have been involved in would be like the design of a manufacturing system. Looking at different alternatives and deciding which one to choose. I worked at Firestone for a while. Looking a process on the floor and looking at different alternatives for manufacturing the product.
Student: How to normally approach the process? Do you have a specific method or do you go by common sense?
Prof: Basically you develop your alternatives. You learn as much as you can about what is going on. I mean, I am an IE, so typically I’m looking at a process and how can we do it better, cheaper, increase production rate, decrease the production cost, whatever the case may be. So you learn as much as you can about it. An IE has lots of things in their toolkit to look at alternatives. We use simulation modeling, mathematical modeling, and basic flow-to-process charts. You develop the alternatives and then compare them. And a typical way to do that is to use computer simulation to model your different alternatives and you have some performance measure that you are interested in, and you see how they compare on that performance measure. Then your decision is based on which one has the best value for the performance measure. Of course, you often are looking at more than one thing at a time. You have multiple criteria, so then you have to come up with some way of weighting the criteria. And then for each alternative, come up with a total score. Then you pick the one that gets the highest total score based on all the different criteria that you might be looking at. And it would be the same thing with a product. You can develop different scores for different products or different designs of the product. You have to put weights on the criteria and score them somehow so that you can compare the different approaches.
Student: Are there any typical problems that you encounter?
Prof: That is a pretty wide-open question. Well, coming up with the right criteria is always a problem. Usually there is more than one person involved and they don’t have the same criteria. And even if they have the same criteria, they don’t put the same weights on the different criteria. You might think that one thing is more important. And I think that it is a criterion, but this one is more important. Those kinds of issues come up.
Student: How many people did you typically have working on this?
Prof: Highly variable. Depends on the situation. Depends on how many bosses are involved, how many supervisors are interested in it, or whatever the case might be.
Student: Do you know of any outstanding references for decision-making?
Prof: There is an organization called the International Council Of Systems Engineering. There is a process called Analytical Hierarchy Process. I think there is a textbook. I don’t know what the name of the book is. It is a way to trade studies. Say that you have five alternatives and ten criteria, how do you do the comparison of those five based on the ten criteria to make a choice? And that could be anything: design of a product, design of system out on the plant floor. Any time you have alternatives and a number of criteria to measure how good each one of those alternatives are, then AHP is to be one process to help you make that comparison so that you can choose between the alternatives. I am sure there is a textbook that talks about it. There is also a paper called “Priority Setting in Complex Problems” by Thomas Saaty. And that is in the IEEE transactions on Engineering Management. August 1983. That gives you an introduction to AHP. There is also a software package that uses AHP called “Expert Choice”. I found that software, but I haven’t learned how to use it yet. But basically it will do the AHP process.
Student: How does it go about using AHP? How does the software work?
Prof: Well, it would take me a while to tell you how AHP works. But it basically does a comparison. You start out with your criteria. Then you build a matrix with the criteria along the vertical and horizontal. Then you rate the criteria against each other. This criterion is way more important than this one. This criterion is somewhat more important than this one. Like that. So you build this little table. And then based on this table you end up with weights that you would use for the criteria. In other words, you get a table that says that this criterion gets a weight of 0.2, this one gets a weight of 0.4 and so forth. Then you look at each alternative. You basically build another matrix comparing the alternatives to the different criteria. Then you decide how well each alternative satisfies each criteria compared to the other alternatives. Then you use the scores from that and the weights you get to finally get a total score. It is a little hard to describe in a nutshell. But basically you are doing comparisons of all the alternatives and how each alternative performs for each different criterion. And once you have done that, then you are able to use the weights that you developed initially to find the total score for each alternative.
Student: Do you feel that software does a real good job with this?
Prof: Well, like I said, I haven’t really used this software yet.
Student: Have you used other software?
Prof: Just spreadsheets. This software should give a lot more capabilities. So, it is probably worthwhile to have. I just haven’t used it yet. So I can’t really speak for it. I can’t really tell you what it does. I do know that it does things over and above the basic AHP.
Student: Have you used any other formalized procedures?
Prof: Even before I used AHP, I did something similar. Just used the idea of coming up with your alternatives and criteria and deciding how each alternative performed under each criterion. And then coming up with weights and a total score. I have done that before. People use that for something as basic as deciding what house to buy or what car to buy or whatever. And I certainly used something like that when I was at Firestone. Or my students do that in my design course. We do real-world problems for companies, and I have had a lot of them apply the AHP process. It seems to work fairly well.
Student: Do conflicts ever arise between using a specific method and common sense?
Prof: Sure. Sometimes you can do something strictly analytical. So you come up with a number, like how fast your car goes. You have a number for that, as opposed to whether you like red better than blue. So one of the conflicts you frequently come up with is some subjective evaluation versus objective evaluation. So one of the problems is which is more important, the objective or subjective value. Because you could certainly get something that objectively points to some value, but when you start looking at it subjectively then there are some good subjective reasons for not making that choice and choosing something else instead. It could be something as simple as where to locate your home office. And you do all sorts of things like transportation; how much is it going to cost us if we are here for travel and so forth. If we are in Omaha, then how much will it cost? Versus Chicago, versus Los Angeles, and so forth. And then you get down to the fact that the boss is from Chicago. And that is a fact of life. He may look at it and see that you would save money if you locate it in Omaha, but he still wants it in Chicago. So certainly things like that come up.
Student: Did you ever have problems like this when you had to present your decisions to managers or your bosses?
Prof: Not personally, but I have seen it. Like my students at the end of the semester presenting to the industrial partner. And the industrial partner might say that the students came up with a pretty good analysis, but there are subjective reasons why that isn’t what we really want to do. And it could just have to do with company policy or something like that. Sometimes it is just how well you present. The best idea in the world isn’t very good if you do a poor job of presenting. How good of a salesman are you?
Student: When you are weighting your decision factors, are there any that you tend to emphasize the most? Or the least?
Prof: That is pretty problem specific. That is different for each decision. But obviously if you are into manufacturing then cost is always going to be one of the most important. Or something cost related. Either decreasing cost or increasing revenues. So that is the bottom-line. Dollars is the most common. There are others things, like quality. But quality is ultimately going to affect revenues. So probably the most important thing in general is dollars, if you are talking about manufacturing. How is it going to affect the bottom line? Obviously there are other things that come into play. Sometimes it comes back to how you weight criteria. When you get to the end and it says that one is better than another one. But we can’t do that one because OSHA says that we can’t. Another problem is that it is just hard to put dollar figures on some things. As an IE and designing a manufacturing system, we have to take into account human factors. Ergonomics. It is pretty hard sometimes to put a dollar amount on how easy it is for a worker to work at the machine. What does it cost me if I don’t make that change? I know that they will probably have more back injuries or whatever. But it would be nice to put a figure on that so that when you present to somebody you can say what the potential savings in health costs are.
Notes from interview with Jo Min (9 November 2001)
Prof: Probably the question is your understanding further into textbook market; probably derive the rest of question and answers.
For example: Suppose your question is commercial type product with competition is intense and a lot of people willing to supply.
In such situation, your behavior is way: if product happen to be the fastest chip that ever been amazing, the number of people that willing to supply these product is limited, meaning you, so in such case that you’re in completely behavior.
So the rest of commercialization will follow the nature of your product.
So, for example, if it’s really the fastest chip, you really don’t have to worry about how much to convince people or how to raise money or how to get better deal than your competitor in terms advance market. In the other hand, of you happen to be making an improvement that is highly competitive to everybody else that’s offering, then you’ll have to focus on marketing, costumer relation management, and general promotional product, your entire focus.
But the question is do you understand your product enough so you know what your product is belong to which group, if you think that this is a highly morrocolistic product, but no body think so, and then you’re in a serious trouble.
Student: Are there any other method that is to use or have been used to go through that process or you decide what your strength is?
Prof: Specifically, you’re talking about methodology like analogy, higher artic process AHP? And Programming and data analysis.
In term of methodology, there are quite a few. But the key question is if the commercial product actually capable to generating, changing process of get to be develop product. That product got to be suffisticated product, you don’t want to use the same decision-making tool that are in the market.
So then, we are talking about gigantic software that we have to use super computer. In that schedule, we have to limited to the few pages.
You can use module A or B out of like 200 modules. So full page developing that kind of system. It’ll be highly sophisticated that belong to certain classes.
Student: Are there any typical problem that everybody have to go through that process in decision-making?
For example: like there’s a team that’s trying to decision about how a product or what’s next step should be or how it should be improved. Some people think different from other, how can we overcome that technical problem?
Prof: There are several approaches; decision-making is such a tedious process because people seem to have different criteria in decision-making. Some people more concern about safety while others concern about prices, yet others like speed of the chip, looking for fastest. It’s depends on the costumer and their background.
Student: How do they go about balancing?
Prof: If it code programming for example, they have specific code regarding code to safe to level and maybe speed and then they try to minimize the way to deviation these specific code. And of course the minimizing totals some of all the deviation. And that’s how they balance.
Student: How many people do you think typically involve in that process like just a small team or a group of committees involve in this thing?
Prof: The important question of that is the degree of expertise, big committees maybe good in a sense that everybody participates. Sometimes, only half of them a reaching on goal whiles other half is mislead. What’s the point of having huge number of people?
For example, we’re deciding whether we use electronic designing a verses electronic design B in circuit board. And only 2 people know about it. It’s not a good idea to have a big group of people. So the important point is how knowledgeable in the committee approach.
Student: Do you know any references such as book or article to search concerning this decision making process?
Prof: In term of decision making process, there’re just too much. In term of commercialization, decision making process I don’t think there’s anything immediately available. You have gone through Technology Management, Engineering Management, and you can go from there to search and find publications about it. Or go to the library talk to reference desk they’ll sure be helpful. Index and abstract databases you can access from your computer.
Student: What are other factors that are important in commercialization?
Prof: Technology, improvement, impact on economy, depends on the criteria where the sources and who you’re talking to.
Student: Are there any strength or weaknesses when you’re talking about AHP, goal programming?
Prof: Absolutely. For example, if you looking at AHP, then you have to look at what’s call paralyze composer, if you happen to be good that those, then it’s okay but a certain people are not really good at it, like A is refer to B, B is refer to C, C is refer to A. So that’s limitation to AHP. Goal programming, you really have to have a goal; you have to know what your goal is otherwise it’s hard to accomplish. But furthermore, there will be a long time; we need a lot of time to improve these effective tools.
Student: So, you don’t think it’s feasible at all of this automate decision making software.
Prof: You really want to comprehensively understand what the scope is. You can design a cloth that will fit everybody, you can have a huge blanket that will fit everybody, but if you want this person to look good, do you think you can design one piece of cloth that fit everybody and still make him or her look good? It’s this possible. It’s obvious that you can design anything, software development, AHP, so and so.
Notes from interview with Van Aucken (8 November 2001)