​Keizo Ishii (President of Quint Co., Ltd.), who was invited as a lecturer at the career path seminar “Creating Markets: Software Development in Unexplored Areas” held on October 11, 2019, gave a lecture on mathematics. We asked students and faculty members who majored in

海外勢に席巻される前に参入してほしい

hello everyone. My name is Ishii from Quint Co., Ltd.

Among design software, I have created software for optimizing structures in particular, and it is being used in the manufacturing industry, including automobiles. As a special example, the research institute of a sporting goods manufacturer used our software to halve the weight of the spikes used by famous major league players over the course of 10 years. I'm here. However, it is mainly used to reduce weight and solve vibration problems when designing fairly solid things such as automobiles, electrical machinery, and steel manufacturing.

Almost 35 years ago, I started a company with the desire to somehow contribute to the Japanese manufacturing industry with the software I created. Japan's CAE (Computer Aided Engineering, software for designing and analyzing products) is currently in a predicament, and it is only a matter of time before it is overwhelmed by overseas companies. The reason for this is that there are very few people who have studied mathematics and physics in Japan.

Many people who are active in software research and development in this field overseas have Ph.D. degrees in mathematics and physics. The difference is emerging today, so if you feel a little bit of interest after listening to my talk today, please enter this field in the future. I want Japan to become even more energetic.

基幹となる技術を持つ

Our main products also challenge topology optimization, shape optimization, and automatically returning the calculated shape to CAD. I'm doing various things, but my company still has only 18 employees. While the current state of CAE software, which is essential for manufacturing in Japan, is being pushed by overseas forces and is in a predicament, how can our software produce results in the world's leading automobile manufacturing industry? Let's talk about Taka.

I founded my current company in 1985. Since that time, CAE software has been mostly shared by large overseas companies, such as Dassault, which makes French fighter planes. This Dassault has a big software company, and in the design of aircraft and automobiles, software called CATIA is sweeping the world's aircraft and automobile industries. Huge conglomerates such as Dassault Systèmes, a subsidiary of Dassault Systèmes, and Siemens PLM Software, which was acquired by Siemens of Germany, have almost dominated the software field of CAE. There are no products made in Japan. In such a situation, I found a useful field in a niche that is not general-purpose, and in a place where a huge software company has not touched.

At that time, I decided to develop a one-of-a-kind CAE software based on the university's unique research as a business model. And when the Japanese manufacturing industry develops products, we created the company with the desire to contribute with this software that we independently developed.

大切だったことは人との繋がり

ジャック・レモン　Dr. Jason R. Lemon

There are three people who were particularly impressed, as if their hearts were gouged out as they talked. One of them is Mr. Jack Lemmon, who founded a company called SDRC in 1967, and started a vibration consulting company, which has grown tremendously. But he was more interested in technology than business. I wasn't interested in business directly, so the company grew bigger and bigger.

In fact, he was the one who coined the universal term “CAE” that we use today. Originally, this person was an assistant professor at the University of Cincinnati, and when he was 32, he founded a company called SDRC. About two years before the word CAE was introduced to the world, I heard various stories from him, and felt that there were so many dreams in the field of CAE, and I wanted to try it someday.

ジョン・ホルキスト　Dr. John O. Hallquist

The second person is the same age as me, and his name is John Holquist. It's to his credit that car companies around the world are now doing pretty well in crash simulation. This person has been researching collisions at the Lawrence Livermore National Laboratory, a national laboratory in the United States, and has created a collision simulation code called DTNA3D. At one point, I encouraged him to go independent, and a few years later, he created a company called LSTC, and the LS-DYNA collision simulation software that was released from there has now taken the world by storm and has an overwhelming market share.

When I first visited him at the Livermore Laboratory in 1982, after we had talked a little, I asked him to come over to my house because I was hungry. We had a blast talking about various things. At the time when I was leaving, I was told to take all of the 2,400-foot storage media with five magnetic tapes under my right arm and several 30-centimeter-long manuals under my left arm. what's in it When I asked about it, I was told that it contained all the source codes of the programs I had developed so far. Since it was made with the government's money, it will be distributed equally, so you can take it home, so I studied.

菊池 昇 先生　Prof. KIKUCHI Noboru

この方は日本人で35歳でミシガン大学の教授になった方ですが、今は日本に滞在していてトヨタグループのトヨタ中央研究所の代表取締役所長をやっています。すごく偉い方なのですけどとっても気さくな方で、僕は彼と38年付き合いがありまして、「俺と石井さんはコインの裏と表だから死ぬときはお互いにキーボードに突っ伏して死のうな」と固い約束をして（笑）います。

What the three of us have in common is that we have a thorough dream and technology. Dreams and technology contribute to the world and the manufacturing industry. And most of them had experience teaching at university, so they went to a company to make use of that experience. That's why all three of us are very solid in basic studies. This is the most important thing.

99.9% of those who do the application first without doing the basics will fail. While talking about such things, I, who was a poor person who was incomparable to them, created the company on the spur of the moment.

３つのこだわり

I have been running the company for 35 years, and I believe that I am here now because of these three commitments.

The first is the dream I got from my predecessors. In the field of CAE, even if we are small in scale, we are not a huge company with thousands of people overseas, but we develop innovative pure Japanese software that can provide ideas to people who use it.

Second, my goal is not to make money. As with the three people mentioned earlier, I wanted to help the manufacturing industry that uses software to develop better products.

The third is challenge, challenge. I stubbornly decided not to imitate others. The more I say this, the more I get myself into trouble. But I don't imitate others. Stick to originality and always try to challenge new things. This is very important. Even if you make a copy of someone else, it will be destroyed in a blink of an eye when a huge capital comes in.

構造最適化は余分な皮下脂肪を取ること

数学に限らず学問を修めることははすごく大変だと思うのですが、自分がバカになれた時に一段上がれます。それから自分がハングリーになった時に追求ができます。これだけは自分の経験としてお伝えしておきたい。（Steve JobsがStanford Universityの卒業式で講演した時に言った有名な言葉です。）

I have been using the finite element method as a discretization method for almost 50 years, and structural optimization for about 35 years as a core technology. As for what structural optimization does, it's easy to do, just drink beer and get rid of the subcutaneous fat in your belly.

When applied to a structure, there are places to hold down and places to apply a load, and when you want to reduce the amount of material a little more in a structure that resists force, shape optimization is a method of scraping the surface. Various methods have been proposed so far, and some people said that shape optimization ended in the 1980s. However, there are people who have continued to research new theories, and now we have technology to remove belly subcutaneous fat more neatly. This is shape optimization.

However, if there is a stoic person who thinks that belly fat is still not enough, what should be done is to remove not only subcutaneous fat but also visceral fat. If you do your best to remove the visceral fat as well, the internal organs are inside the body, so there will be holes there. This is the ultimate optimization, and the mass is greatly reduced.

However, there is only one drawback to this, and a problem arises in the manufacturing process of how to make something with holes in the structure like this. In recent years, 3D printers have solved this problem.

By the way, when and where did the technology of continuum topology optimization come about? The paper was first published in 1988. Martin Benso, a professor of mathematics at the Technical University of Denmark, When Professor Noboru Kikuchi of the University of Michigan was staying at the Technical University of Denmark on a sabbatical, the two of us discussed and created the concept of continuum topology optimization.

Even after 30 years, this paper is still accessed from all over the world, and it is ranked in the top 3 in the ranking of access among Japanese. What makes this paper so unique is that when trying to remove the aforementioned visceral fat or subcutaneous fat, even if the target structure is filled with material, it can be used as a porous body with many small holes inside. I assumed. The reason why they assumed a porous material was that they already knew the homogenization method studied by Lions et al. of France in the field of applied mathematics at that time. I listened directly to a lecture on the homogenization method from the late Professor Jacques-Louis Lyons of France, and was wondering if it could be applied to engineering, and what I came up with was this topology optimization.

Sometimes it is convenient to assume a porous body, and when this is optimized, the design variable becomes the pore size of this porous body. Since the space is cyclic, the minimum unit of the periodic structure in a certain area is the unit cell, and the size of the hole in that cell is adjusted. When becomes small enough, it becomes a full material state. Interpolate continuously in between. When you finally try this, for example, the upper shape will come out as the lower shape.

Thirty years later, topology optimization is flourishing all over the world, and there is a world conference on structural optimization called WCSMO (World Congress on Structural and Multidisciplinary Optimization). It is held once a year, and at the first conference, there were only a few people who made presentations under the title of Topology Optimization. It is topology optimization. It's been 30 years and it's getting more and more popular.

I made software based on the first paper at that time. The paper was published in 1988, and the commercial software OPTISHAPE was released the following year. I went around various companies, including automobile companies, and asked them to use it. More than ten years later, a bookstore called Springer published a thick technical book called Topology Optimization. There was a chapter called Software for Topology Optimization, and it was written that in 1989, a Japanese company called "Quint" released commercial software under the name OPTISHAPE. We were very happy with the proof that we were number one.

Continued on (2)​