First-Hand:Optimizing the Project Engineering Process

From ETHW

Submitted by C. R. Schmidt

An employer will almost never transfer a senior man to a new product line involving new technology. In this situation, the employer invariably chooses a junior or novice engineer. His reasoning is simple, both the senior and junior man will take about the same time to do the job and the junior man gets much less pay. Hence, there is an obvious cost saving. No age discrimination here, just basic economics.

Engineers of all ages should look toward retraining themselves as new technologies emerge. Companies-particularly the smaller ones which employ the bulk of engineers-are more fragile than they appear. Sometimes they become victims of unimaginative marketing; sometimes they are bought and restructured; sometimes the original owners die. In any event, companies do go out of business. Now a pension bound engineer who has not kept abreast of some new technology can be in big trouble. This disappearance of jobs probably accounts for the largest disappearance of engineers from active practice.

An engineer's retraining begins with oneself. The emerging technology which appeals to him or her should be the first consideration. Without a special interest, nothing will happen.

Out of my interest in writing programs for the Intel 80C48, I acquired a Commodore VIC- 20 at a price of sixty-nine dollars, a twelve inch black and white television at sixty-five dollars, a Promqueen at two hundred dollars, and an OKI development board for simulating an 80Q48 at about fifty dollars. I set myself the goal of writing a serial keyboard routine that would minimize the number of In-Out ports required of the 80C48.

It took me two months to get the hardware together and another two months to assemble it and write the program. The day finally arrived when I could press any of sixteen keys and have its symbol come up on a liquid crystal display, a fairly conventional result. However, I had accomplished this with only three In-Out ports of the 80C48 instead of the conventional eight ports. Well, even though I had retrained myself in this field, my company brought in a junior man when they decided to go into microprocessor designs. This young man, twenty-seven years old, got his experience at Eastman Kodak in Rochester. In a few more years, he will have to think about retraining himself.

As far as I am concerned, I know I am competent in the microprocessor field. When the time comes I will prove it. After all, at the time I was only sixty-six years old.

An engineer's first goal is to design a product which is not only useful but also makes money. As far as permanence is concerned, no matter how original an engineer's design may be it will never last forever. The inexorable progress of technology sees to it that engineering designs will find it hard to be in continuous production for thirty to forty years. So the practice of engineering may not achieve permanent results, but it certainly is an absorbing adventure.

The scenario today begins with the marketer (usually with management's blessing) presenting the engineering group with the customer's wish list for the product and suggesting the direction the product design should take. The engineering segment, after commenting adversely on the barrenness of marketing's design suggestion, then presents a menu of product benefits versus estimated cost per benefit.

The marketing people usually don't know exactly what the customer wants. The problem of obtaining the customer's true need involves making him or her understand that little-used features will add to the product's cost without making it substantially more useful. Once the customer's thinking is refined to consider an optimum design, his or her true requirements will emerge.

A design engineer who can think in terms of hardware required to provide these benefits can be very helpful at this stage. In 1986, it was suggested by an engineering vice president at General Electric that a design engineer accompany a sales or marketing person to call on the customer for the purpose of asking the right questions. Successful designs result when the product requirements are optimized. This is best achieved when one person makes the tradeoffs.

This is because the committee approach is doomed to failure from the start. The trouble with design committees are that the product design becomes a compromise in which each participant tolerates the others' ideas, so that all pet ideas are included somehow. This is far from an optimization process.

The complete design should be made by one knowledgeable engineer. If his or her design is not acceptable, he or she is the wrong person for the job and another engineer should be selected. The design engineer must not retreat to an ivory tower in the design process.

On the contrary, he or she should try to include all the suggestions presented and through discussion of individual requirements with others clarify all the possibilities in his or her own mind. The process should include the maximum number of significant benefits with the minimum of hardware and cost. It can be a very engaging process especially when it's on the right track. In this process, the engineer can have no clever or cute pet ideas that he or she is stuck on. The optimum design must include the best, most suitable ideas, whether they're the design engineer's or somebody else's. The end product must be considered and visualized completely in the designer's imagination before the final specifications and sketches are made. In some cases, working prototypes may have to be made to prove feasibility or suitability of some functions. Once these determinations have been made no additional detail designing should be undertaken until the complete product design has been accepted.

If there are competing approaches to the product design, each should be considered thoroughly and objectively. Only then can the selection from the alternative designs be made. The best design is then presented to sales, marketing and management for acceptance. If the job has been done properly the prospect for acceptance will be quite bright.

The constraints of meeting the target design specifications and target price goals are commonly understood by project engineers. The constraints of project expense and date of completion of the product design are grudgingly recognized or not recognized at all by many project engineers. This is particularly true if the project expense budget is low and the time to completion is short. Engineering design time is a company investment and as such cannot be squandered. So if the budget is very tight, the engineer may be obliged to restrict himself or herself to a choice of known technology or possibly to just an extension of the company's existing product line. No examination of emerging technology may be possible. This obviously will not result in one of the company's future breadwinners, but a return on investment has to be anticipated at some point. However, if company management regularly short budgets its products, it will be condemning the company to new product offerings not keeping up with the new technology. Consequently, business volume will decline.

Management must recognize the talented engineer and assign him or her to short budget projects. Chances are, the talented engineer will accept the challenge and produce a beneficial result. However, there is no way even the most talented engineer can overcome a severely short time and money budget. Management is responsible for this phase of the project. There is no free lunch in the free enterprise system. If management cannot see a substantial return on its investment with a new product consistent with its established quality level, the project should never begin. An engineering bail out is a long shot maneuver undertaken only by start-up companies and companies about to go out of business.

Regardless of the circumstances, the engineer will make the compromises needed to meet the price and performance specifications in the time and money limits allotted for the project. This type of performance is what has made engineers worth the money they are paid today and, indirectly, is what has brought about their relatively high starting salaries.

In many successful projects, I have felt, after the money starts to be made, that I had only a small part in the outcome and that all the others had done much more than I. Perhaps this feeling is the result of another job related attitude that I have.

I listen to the members of my team, whether they are technicians, design draftsmen, or purchasing agents. In discussing the various aspects of the project with them, I encourage them to make suggestions and observations. You encourage people when you don't harshly criticize their offerings and when you entertain their suggestions to a conclusion.

Although I was not the project engineer at the time, the following example illustrates this point. In the late sixties, we found that the electronic parts of our instrument designs were much lower than the cost of the instrument enclosures-due mainly to lower semiconductor and power supply costs.

Together, with the corporate industrial design group, we cast about for alternatives to the sheet metal fabricated construction we were using at the time. We explored the possibility of extruded aluminum side sections with mitered "V" grooves as fold points to make up the box together with a sheet aluminum bottom captured by extruded grooves. This was attractive because extrusion die costs were low. We even explored sand castings, one of which won a design institute prize at a national instrument show. Our problem was that our new design sales quantity seldom exceeded one hundred units a year. So we were trapped between high fabrication costs and high tooling costs. In the course of our discussions with the corporate industrial designer, he mentioned that "if the annual instrument volume is greater than one thousand a year, the California designers consider die castings." Those were magic words.

At our operation, we had an ancient line of slide back conductivity instruments with fabricated cases that were long overdue for a redesign. The combined annual volume at that time for these instruments was eighteen hundred pieces. The proposal was made that we make a universal case design that could be used for both the old slide back and new meter type instruments.

The old enclosure costs were about twenty-eight dollars a piece, the new die cast case cost would be about two dollars. This would bring the enclosure cost of new designs in line with the lower electronics costs. The return on investment in the die casting was several hundred percent in the next five years well above the fifty percent required by corporate accounting.

This design process-in which a part to be used in new production is also used to improve the profitability of old production-is called contextual engineering. It is just one way to restore yesterday's breadwinner products to good profitability. Listening for the magic words is a habit that can produce very beneficial results.