Customer Reviews

Axiomatic Design: Advances and Applications (The Oxford Series on Advanced Manufacturing)

5 star:		(0)
4 star:		(1)
3 star:		(0)
2 star:		(1)
1 star:		(1)

 

 

One of great emerging ideas in the past 20 years is that there are tools to help in the design of robust products and systems. There are other approaches: TRIZ, QFD and other heuristic-based methods. There are also references to Dr. Taguchi's robust design methods (DFSS). The point is that the author here has made his point - DFSS as applied to multi-requirement products and systems... and that includes just about every product/system we use and need... has an overlooked flaw. Read this book to understand what that flaw is and how to address it.

Is the material here ambitious and audacious? Well... yes, wouldn't it be if the author were on to something? I think that the key here is to understand where this information can be immediately used in engineering practice. The principle of design decoupling and Okcam's Razor are not particularly new ideas, but the way that the author is approaching the subject is important. He is trying to get a handle on something that has been to date very heuristically practiced, and often not well executed. If one understands the principles of, for instance, software engineering, it doesn't take long to understand where the author is going with his subject. Suh is a mechanical engineer (MIT), and this fact, in itself, is very surprizing and encouraging. The axioms that Suh present are necessary in order to achieve some sort of order to the discussion. I personally have managed to get past the initial objections to his reuse of certain terms and emphasis on certain axioms. For instance, Suh's principle of "minimum information content" is actually a statement about "less is more" in design from a reliability and design-robustness perspective.

Now for potential buyers: understand that there are applications of this method and theory that are extremely powerful and effective. This method has great implications to the field of systems engineering - it ties the SE discipline to something that you can get your hands on. In particular, look to the application of the method to the design of physical systems and their corresponding manufacturing systems. This isn't a trivial or invalid subject.

People who need ways to handle complexity and the total system design problem will benefit as will designers of common products. Software and hardware designers can both benefit, although the author uses language that is more recognizable in the realm of mechanical engineering and manufacturing systems. This material will take some time to integrate into your own storehouse of knowledge - so don't rush it. Approach this material with a need in hand, but with the understanding that it will likely modify your view of your discipline. Then you will make use of this material. Note the bibliographies; dig into the applications of this theory. You won't be disappointed in the outcome.

I bought this book for a readings class this semester. I don't have a suitable vocabulary to describe how poor it is. The text is inconsistent and full of errors. The author blathers on without concluding much. He references his previous work and that of PhD theses (which I can only assume are his own students). I question the underlying assumptions of his theory but it is never fleshed out in sufficient detail to be understood and examined.

Some examples:

Page 8, the author discusses history, "there were no exceptions or counterexamples (to Newton's laws) until Einstein advanced the theory of relativity". Really? How about Maxwell's equations, the Michelson-Morley experiments, the theory of the ether, Mercury's orbit around the Sun?

Page 18, The author uses matrix algebra notation and operations on nonlinear equations. Superposition does not apply to nonlinear equations.

Example 1-13, consists of a redesign problem where one constraint is "no increase in cost", and the solution to the example problem is to add a component to the existing design. And that component is free?

Title of one section: "Reduction of uncertainty: Conversion of a design with time-dependent combinatorial complexity to a design with time-dependent periodic complexity".

It goes on. I made it about 100 pages into it.

Is design a science? Or, can design be a science?
The author tries to conclude the principles for "good designs" into two axioms, then use them as the scientific approach to conduct the design activities. It is ambitious and audacious. Unfortunately, this is a false science. The paradox of the author's intent lies on the fact that science, by definition, should be repeatable and universal. However, regardless of the two controversial "axioms", which have been refuted in many literature among the design community, they themselve do not guarantee that different people will arrive on the same design. They are, at most, two design principles that may not be 100% true, depending on the cases.
The domain and tasks of design, is too broad and versatile to be abstracted by any "axioms", 'cause design is a mental process of creativity, and so far no one can sucessfully describe creativity in scientific terms.
Therefore, the answer is "no" to the question in the beginning of the review, at least not for this book.