Catálogo de publicaciones - libros

This paper discusses design issues that are equally important for engineers engaged in system development as well as for researchers interested in design theory, in particular, in innovative design synthesis. The paper aims to formulate unsolved design problems, which could be treated better with a help of intelligent computing in general.

Starting with the provisional definition of good design and an innovative design example, a process of new product development is analyzed and necessary activities before and after the design & development (D&D) work are identified. Strategic marketing and planned innovation are important before D&D work, while avoiding wrong products and achieving high system performance/ qualities are critical after it.

After advocating the importance of concept design at an early designstage, the author introduces, first of all, conventional design paradigms such as analysis-design-evaluate (ADE) paradigm, case-based paradigm and artificial intelligent (AI) paradigms and the like. As far as design of highly reliable systems with high performance are concerned, a new design paradigm, where common sense of design is changeable at any moment, is needed and identified by the author.

The practical and effective design theory focusing on conceptual design, unfortunately, is not readily available, practicing engineers, in general, are conducting various design works quite empirically. The most promising approach, for a concept design level in particular, could be explored by learning lessons from various perspectives including;

After discussing the above items from 1. to 4. with each example, the issue related to 5., design theories, are explained in detail. First, design basics such as (1) definition of design, (2) nature of design, (3) meaning of design, (4) act of design, (5) classification of design, (6) process of design, and (7) method of design, are introduced.

Then, the author has selected and evaluated three approaches, for every designer in mind, to be effective and usable from the system developers point of view. These are Principles of Design (Sue), Sciences of the Artificial (Simon), and Universal Principles of Design (Liddwell et.al). In addition to these useful principles identified, Is Is-not Analysis is conducted to define a concept of a good design engineer.

Finally, unsolved design problems and related research themes are formulated as a summary statement. Researchers are most welcome to penetrate into these quite difficult areas of design synthesis, which, if solved or enhanced by applying intelligent computing, could revolutionize the design practices in engineering for today and tomorrow.

Following the overview of Interactive Evolutionary Computation (IEC) research, three recent topics are introduced: micro-electrical mechanical systems (MEMS) design with IEC, psychological measurement with IEC, and extended IEC with physiological feedback. IEC is an optimization system that optimizes a target system based on human subjective evaluation. There are many systems whose performances for system optimization are not measurable but can be evaluated by humans when we hear, see, or touch the system outputs. IEC is a technology that is applicable to such systems. It is widely used now and some of its applications include generating computer graphics (CG), music, or artistic design, designing acoustic or image processing filters, controlling virtual reality or robotics, designing in engineering, and others. First, we review this status-quo of IEC research.

There are many links among electric appliances. We live with personalized appliances, which are achieved by links. AIBO(SONY: http://www.sony.co.jp) is famous as a representative example of personalized robots. Also, personalized cleaners are becoming popular. In the meantime, IT made remarkable progress especially in personalizing-related technology; authentication technology, animation on web page, and web camera. Collaboration with Network Technology is popular in RT field. Users who are not familiar with computers can use IT through friendly personal robots. It is expected wide diffusion of IT. RT, AT and IT are growing in spiral by interacting one another as .

Many efforts have been being made as in . For example, Ministry of Public Management, Home Affairs, Posts and Telecommunications held workshop headed by Prof. Tokuda for networked robots technology, which aims to build an open platform. On the other hand, the policy of developments of robotics depends on each group or laboratory, but there is a movement that carries on the standardization. We think Ontology is necessary for Natural Interface and mutual understanding between different machines such as robots and cars.

“Can we predict the future? Or, should we produce the future?” This is coming to be an essential question in business domains. In Chance Discovery, the focus is on a chance meaning an event or a situation significant for making a decision in a complex environment. Interdisciplinary discussions by researchers from philosophy, sociology, artificial intelligence, finance, complex systems, medical science, etc, contributed to the methodologies of Chance Discovery since it was initiated in 2000. In this talk, methods and applications of chance discovery are presented. The applications include earthquake prediction, diagnosis of hepatitis patients, marketing scenario making, and product development. In the presented process of chance discovery, results of visual data mining for data on the external environment and the thoughts of user about scenarios in the future are integrated. This urges trustworthy scenarios to emerge. Some scenarios work as predictions of the future trends and some as production of values in the future. The scenario-emergence effect from the methods of chance discovery presents a general clue for answering the question.

In recent years, much discussion of “Complex systems” has been made in various research fields. As my main research area is the complex/harmonious systems engineering, I have, on many occasions, been asked to explain it by those who are interested in this area. When encountering such a seemingly straightforward question, it is not always easy to give a simple answer.

Although some indicate that “It is important to define the basic concepts, however, it could turn out to be just playing with words if not done properly”, it is still important to consider the philosophy behind the complex/harmonious systems. Thus in this paper, the main ideas of the complex and harmonious systems are summarized based on the past literature, and then the concept of the complex/harmonious systems is clarified based on the discussion of systems from the general systems theory.

It has been ten years since the Laboratory of The Harmonious Systems Engineering at the Graduate School of Engineering of Hokkaido University was established. Activities of the Laboratory show how the complex/harmonious systems engineering can apply to solve for problems in various fields. The following topics will be presented:

Visualization of multivariate data is one of the key technologies in the fields of data-mining, kansei engineering, chance discovery, etc. This talk summarized our recent study on visualization methods that could relate data to words.

In this presentation we will discuss potential of biomimetics: how engineers can transform ideas coming from better understanding of living organisms, generated mainly by life scientists, for designing better, smarter systems (or devices) for solving everyday problems.

In the first part of our talk we will define data mining, metadata mining, and knowledge discovery process. Next, we will talk about our data mining projects in the areas of biology and medicine. Then we will show how what we have learned from theses projects, namely, how information is processed and modified in living organisms inspired us to develop new learning mechanisms. In the second part of the talk we will describe how these learning mechanisms are used for solving clustering and medical image processing problems.

Peptides binding to MHC molecules can be presented to T-cell receptor and then trigger an immune response. Prediction of peptides binding a specific major histocompatibility complex has great significance for immunology research and vaccine design. According to their different structures and functions, MHC molecules can be classified into two types. Most of early studies often focus on MHC class I, but seldom on MHC class II. In this paper, we present a method for MHC class II binding peptides prediction using Fourier analysis and support vector machines (SVM), the novel prediction technique is found to be comparable with the best software currently available.

The presented study comprised the employement of a neural network (NN) algorithm, radial basis function (RBF), for the purpose of daily trip flow forecasting in Istanbul Metropolitan Area. The RBF NN predictions were quite close to the observations as reflected in the selected performance criteria.

Previous research into the neural classification of E.coli promoters has focused on the use of raw DNA sequences and alignment methods. In this paper, we use sequence dependent structural profiles of DNA to train neural networks for promoter recognition. In addition to this, we evaluate the impact of different types of non-promoters used in training and testing on the classification accuracy. 872 E.coli promoters were used in addition to three types of non-promoters, random sequences with the same base frequency as the promoter sequences, genes selected from E.coli and random sequences with the same base frequencies as the gene non-promoters. Raw DNA sequences were then converted to stacking energy and GC-trinucleotide profiles. We found the promoter classification accuracy using structural profiles was comparable to other methods. However, our approach has the advantage of not requiring finding the -35 and -10 hexamers and alignment of the DNA. Overall, using non-promoters from coding regions and random sequences with the same base frequency as the gene non-promoter resulted in the best classification accuracy.