Matriz qfd
Páginas: 44 (10816 palabras)
Publicado: 4 de junio de 2010
ullman-38162
book
June 20, 2002
11:32
C
H
6
A P T
E
R
Understanding the Problem and the Development of Engineering Specifications
6.1 INTRODUCTION
The Mariner IV satellite was to be packaged in a rocket, its solar panels folded against its sides. After launch the satellite was to spin so that the solar panels would unfold by centrifugal force and be locked in astraight-out position. Because these panels were quite large and very fragile, there was concern that they would be damaged when they hit the stops that determined their final position. To address this problem, the major aerospace firm that had the Mariner contract initiated a design project to develop a retarder (dampener) to gently slow the motion of the panels as they reached their final position. Theconstraints on the retarders were quite demanding: they would have to work in the vacuum and cold of space, work with great reliability, and not leak, since any foreign substance on the panels would harm their capability of capturing the sun’s energy during the satellite’s nine-month mission to Mars. Millions of dollars and thousands of hours were spent to design these retarders, yet afterextensive design work, testing, and simulation, no acceptable devices evolved. With time running out, the design team ran a computer simulation of what would happen if the retarders failed completely; to the team’s amazement, the simulation showed that the panels would be safely deployed without any dampening at all. In the end, they realized that there was no need for retarders, and Mariner IVsuccessfully went to Mars without them. This story is an example of how a lot of time and money can be wasted designing the wrong product. Surveys show that poor product definition is a factor in 80 percent of all time-to-market delays. Further, getting a product to market late is more costly to a company than being over cost or having less than optimal performance. Finding the “right” problem to be solvedmay seem a simple task; unfortunately, often it is not.
111
ullman-38162
book
June 20, 2002
11:32
112
CHAPTER 6
Understanding the Problem and the Development of Engineering Specifications
Besides finding the right problem to solve, an even more difficult and expensive problem for most companies is what is often called “creeping specifications.” Creeping specifications changeduring the design process. It is estimated that fully 35 percent of all product development delays are directly caused by such changes. There are three factors that cause creeping specifications. First, as the design process progresses, more is learned about the product and so more features can be added. Second, since design takes time, new technologies and competitive products become availableduring the design process. It is a difficult decision whether to ignore these, incorporate them (i.e., change the specifications), or start all over (i.e., decide that the new developments have eliminated the market for what you are designing). Third, since design requires decision making, any specification change causes a readdressing of all the decisions dependent on that specification. Even a seeminglysimple specification change can cause redesign of virtually the whole product. The point is that when specification changes become necessary, they should be done in a controlled and informed manner. The importance of the early phases of the design process has been repeatedly emphasized. As pointed out in Chap. 1, careful requirements development is a key feature of concurrent engineering. In thischapter the focus is on understanding the problem that is to be solved. The ability to write a good set of engineering specifications is proof that the design team understands the problem. The methodology developed in this chapter for finding requirements helps
I
I
I
Ensure that the requirements are discriminatory. Requirements need to reveal the differences between alternatives to help...
book
June 20, 2002
11:32
C
H
6
A P T
E
R
Understanding the Problem and the Development of Engineering Specifications
6.1 INTRODUCTION
The Mariner IV satellite was to be packaged in a rocket, its solar panels folded against its sides. After launch the satellite was to spin so that the solar panels would unfold by centrifugal force and be locked in astraight-out position. Because these panels were quite large and very fragile, there was concern that they would be damaged when they hit the stops that determined their final position. To address this problem, the major aerospace firm that had the Mariner contract initiated a design project to develop a retarder (dampener) to gently slow the motion of the panels as they reached their final position. Theconstraints on the retarders were quite demanding: they would have to work in the vacuum and cold of space, work with great reliability, and not leak, since any foreign substance on the panels would harm their capability of capturing the sun’s energy during the satellite’s nine-month mission to Mars. Millions of dollars and thousands of hours were spent to design these retarders, yet afterextensive design work, testing, and simulation, no acceptable devices evolved. With time running out, the design team ran a computer simulation of what would happen if the retarders failed completely; to the team’s amazement, the simulation showed that the panels would be safely deployed without any dampening at all. In the end, they realized that there was no need for retarders, and Mariner IVsuccessfully went to Mars without them. This story is an example of how a lot of time and money can be wasted designing the wrong product. Surveys show that poor product definition is a factor in 80 percent of all time-to-market delays. Further, getting a product to market late is more costly to a company than being over cost or having less than optimal performance. Finding the “right” problem to be solvedmay seem a simple task; unfortunately, often it is not.
111
ullman-38162
book
June 20, 2002
11:32
112
CHAPTER 6
Understanding the Problem and the Development of Engineering Specifications
Besides finding the right problem to solve, an even more difficult and expensive problem for most companies is what is often called “creeping specifications.” Creeping specifications changeduring the design process. It is estimated that fully 35 percent of all product development delays are directly caused by such changes. There are three factors that cause creeping specifications. First, as the design process progresses, more is learned about the product and so more features can be added. Second, since design takes time, new technologies and competitive products become availableduring the design process. It is a difficult decision whether to ignore these, incorporate them (i.e., change the specifications), or start all over (i.e., decide that the new developments have eliminated the market for what you are designing). Third, since design requires decision making, any specification change causes a readdressing of all the decisions dependent on that specification. Even a seeminglysimple specification change can cause redesign of virtually the whole product. The point is that when specification changes become necessary, they should be done in a controlled and informed manner. The importance of the early phases of the design process has been repeatedly emphasized. As pointed out in Chap. 1, careful requirements development is a key feature of concurrent engineering. In thischapter the focus is on understanding the problem that is to be solved. The ability to write a good set of engineering specifications is proof that the design team understands the problem. The methodology developed in this chapter for finding requirements helps
I
I
I
Ensure that the requirements are discriminatory. Requirements need to reveal the differences between alternatives to help...
Leer documento completo
Regístrate para leer el documento completo.