Multi-objective Evolutionary Optimisation for Product Design and Manufacturing

Citation

Mario Salmon, I. (2012), "Multi-objective Evolutionary Optimisation for Product Design and Manufacturing", Assembly Automation, Vol. 32 No. 4. https://doi.org/10.1108/aa.2012.03332daa.009

Multi-objective Evolutionary Optimisation for Product Design and Manufacturing

Article Type: Book review From: Assembly Automation, Volume 32, Issue 4

Lihui Wang, Amos H.C. Ng, Kalyanmoy Deb, L. Wang, A.H.C. Ng and K. Deb,Springer,2011,502 pp.,US$229,ISBN: 978-0-85729-617-7,Web Link: www.springer.com/engineering/production+engineering/book/978-0-85729-617-7

This large and very well edited book deals with the typical activities encountered in any design of new products, product families, industrial processes and services: the parameters definition taking into account a plurality of conflicting objectives. In fact, multi-objective optimisation is the key phase in any new design. The goal of this activity is to define a set, called Pareto-optimal, of design parameters showing the relation between conflicting objectives. This multi-objective optimisation (MOO) is also known as multi-disciplinary design optimization (MDO). The book deals with optimisation made by using an evolutionary genetic algorithm.

The first part of the book deals with the theory of the method, plus a new consideration of an innovative application called “innovisation”,while the second and much larger part presents application cases. The cases deal with a very large – over a dozen – and very well diversified application examples, such as: machining, job-shop management, goods distribution network design and management, spring design, robot family planning, and rapid prototyping, all collected by worldwide researchers from the UK, the USA, Singapore, India, People’ Republic of China, and Canada.

Application examples, the core of the book, encompass dynamic systems, too, as a supply chain design and management optimisation, giving rise to interesting considerations on the trade-offs between inventory and batch size.

All the cases have been approached from a rather theoretical point of view and, in many cases, are rather far from real industrial engineering applications. The cases have been selected to show the wide application field of the method as well as the possibility of being applied to completely different types of problems ranging from product design to supply chain management. Alternatively, well known optimisation methods are quoted showing the strength of the genetic approach in “non derivative” cases that are the large majority of real problems.

The book is intended for researchers who are already experts in the theory and application of genetic algorithms and MOO, and requires careful and not-easy reading.

In spite of the research approach, the book may be useful for any industrial “designer” and decision maker, because it shows how multiobject optimisation may be applied in a wide variety of interesting cases, how optimisation has to be defined, and how genetic algorithmic applications work in many engineering cases.

The book has a large number, approximately 200, figures and tables that allow a non-specialist to understand the power and the results of the proposed method. A large reference list of about 500 is given, allowing researchers to deepen their specific knowledge of this field.

In conclusion, this book is for a specialist, but is stimulating reading for all persons having to make decisions in the presence of conflicting objectives; say multiobjective, where there is no single optimal solution (i.e. the optimum for all objectives), but a set of solutions, known as Pareto-optimal, where one selects the design parameter.

Ing. Mario SalmonBologna, Italy