“I want people to discover themes that humans cannot solve alone”: Expectations for WIRED COMMON GROUND CHALLENGE by roboticist Takeo Kanade

──Could you tell us how Ms. Sanji and Mr. Kanade came to work together at the Technology Laboratory?

Shinichiro Sanji (hereafter, Sanji) 　Originally, when I started a global robot competition, I received advice from Professor Kanade. What we discussed there was how to create an industrial ecosystem centered on robots and how to bring about innovation.

It was impressive to hear that there are two axes in how to create an ecosystem. The first is the direction of creating an industrial ecosystem centered on regions and locations. This is exactly the type of social implementation of various technologies around Carnegie Mellon University where Professor Kanade was. Technology was implemented in society by activating the connections between people who were in the place.

The second type is the type in which evolution begins in a ripple effect when core technology is born, and innovation occurs when people compete with each other. An example of this is a project that came out of DARPA (Defense Advanced Research Projects Agency). The friendly rivalry of the gathered people creates innovation. Regardless of the type, I keenly felt that it would be difficult to create an ecosystem without people-to-people connections and places.

So when PwC Consulting decided to launch the Technology Laboratory as a base for industry-government-academia collaboration, we wanted to provide a place where people could gather and compete, and a space where the ecosystem could be put into practice. I would like to ask Mr. Kanade for advice.

──Mr. Kanade himself has been involved in industry-government-academia initiatives for a long time. Looking back, how do you see the recent changes in research?

Takeo Kanade (hereafter, Kanade) 　From the perspective of those in the research and development field, the biggest difference from the past is the so-called IT technology that transmits information and uses it. . I wrote my doctoral dissertation in 1973, almost half a century ago. Since then, the progress of information communication has made the transmission of knowledge instantaneous. In the past, not knowing was the biggest handicap, but now it's completely gone.

In other words, how researchers can use the power of information to travel quickly has become important. Information is all about who is doing what and what is important now. In short, I feel that technology has exponentially increased our opportunities and our ability to realize them.

In my computer vision research, I often use the term perfect storm. A perfect storm is a large storm created by a combination of certain weather conditions. If you look at this in a positive way, you can use it for the meaning of a thousand chances. Today's information and communication technology provides a tremendous amount of power to our ideas by arranging various conditions.

TAKEO KANADE｜TAKEO KANADEWaitakar Memorial University Professor at Carnegie Mellon University in the United States and Visiting Distinguished Professor at Kyoto University Institute for Advanced Study. In 1973, he received his doctorate from Kyoto University and worked as an assistant professor.In 1980, he transferred to Carnegie Mellon University, where he served as director of the Robotics Research Institute (1992-2001). Major research achievements include the world's first facial image recognition (1972), the Lucas-Kanade method, a basic algorithm for video processing (81), the first self-driving car that crossed the American continent "Navlab 5" (95), There is a multi-camera 3D replay system "Eye Vision" (2001) that synchronizes 33 robot cameras adopted in the 35th Super Bowl in real time.

Chemical Changes in Academia and Consulting

──Mr. Sanji, what kind of efforts have you made with academia?

Sanji: Conversely, I myself am a person who has only done industry-academia collaboration. Even before joining PwC Consulting, I have been working to foster the momentum of innovation through discussions with various researchers in national committees, etc., and I have participated in national projects to stimulate research and development. I also tried. In fact, I have experienced manufacturing in the form of joint research.

There may be few people who want to do manufacturing at a consulting firm, but I think it is necessary to work from the concept stage to implementation. The best way to implement this is in the form of a DARPA project, which can be said to be the epitome of industry-academia collaboration.

Also, when creating an ecosystem with academia as a consultant, I think it's important to have hands-on, hands-on experience. Even if you make various proposals in consulting, there are many things that you cannot understand unless you see the actual product. For example, the difference in wireless and wired connections alone creates delays, and differences in the experience of XR and VR. If you don't experience it while actually collaborating, you won't know if the technology can really be used in the field.

On the other hand, people in academia sometimes have a hard time imagining how their products will be used in society. That is why I am keenly aware on a daily basis that it is important for us to create use cases ourselves and for academia to advance the research results themselves.

Shinichiro Sanji | SHINICHIRO SANJIPwC Consulting LLC Partner / Director of Technology Laboratory. After working at a Japanese think tank and a consulting firm, he assumed his current position. He has strengths in consulting that makes use of the characteristics of industry, government, and academia. Engaged in planning for social implementation, consortium launch support, technology strategy formulation, technology roadmap formulation support consulting. To solve social issues through co-creation between industry, government, and academia, we support government offices, private companies, and research institutes from visioning to execution using cutting-edge technology. We provide a wide range of consulting from a high perspective, such as policy planning support, technology evaluation of research institutes, and support for new business execution of companies.

──How common are initiatives like yours, which combine academic knowledge with business implementation?

Kanade: What I find interesting about the United States is that such a concept is common. For example, DARPA, which Mr. Miharu mentioned, has a summer camp for coming up with ideas, where leading researchers and DARPA program managers run various programs for a week in Cape Cod, New Hampshire. A lot of former astronauts, former Air Force generals, former business people, etc. come there and give specific advice. I think it's really great.

I often say that "amateur" is important. However, amateurs do not mean that they do not know things. Professionals are always caught up in many things. The great thing about the business professionals I know is that they speak their minds without hesitation as amateurs. In fact, it is a special ability in itself.

The Technology Laboratory is also equipped with a hemispherical imaging system, and is intended not only to develop technology, but also to function as a base for experiencing it.

Different viewpoints create innovation

──Mr. Kanade, you are a so-called "amateur", and have you ever had an experience that opened up your research horizons? do you have.

Kanade: There are so many of them. When Ichiro is on defense, when he throws it at the catcher, it hits the catcher's mitt perfectly. In theory, this happens because you are calculating and throwing at the correct speed and the correct launch angle. So, the experts break down the process into several steps, measure the distance from the position of the fielder to the catcher, know the wind, and then calculate with Newton's law. But if you think about it, Ichiro wouldn't have done that.

Perhaps, but somehow you get the final result by bypassing the notion of distance or speed. But the correct answer is that you don't need intermediate results in the technical sense to get the final result. On the contrary, if you try to obtain it, problems will arise. If you can execute each step perfectly and measure the distance and wind speed correctly, you should be able to find the correct answer using it. However, the answer is not correct. The reason is that the data are not precise enough. In other words, if you openly try to decide the answer at each step, errors will enter somewhere, and it will work until the end.

Current deep learning has the implication that the process is not decomposed in the middle from the beginning to the end. In the field of image analysis, there are surprisingly many successes with this method. Science is made by taking things apart, so of course it's good to take things apart, but there are times when it's better not to take things apart.

──As a scientist, how do you think you can create a common basis of understanding with various people, such as engineers and consultants?

Kaneide: I think that the basics are nothing but talking. If you don't notice yourself, ideas won't come out. However, there are a lot of people who keep their thoughts a secret, probably because they have a strong feeling that they will be stolen if they say it. I think people who tell their ideas to others are smart. Ideas are sharpened and better by telling people. I understand that the award, which can be said to be an attempt to compete by presenting one's ideas to others, is also such an attempt.

Sanji: I also think that expressing an idea is a means for the person himself to confirm its content. I can't convince myself when I'm in agony by myself, but when I explain it to others, I brush up on my ideas. With more descriptive power comes a story.

Furthermore, I think it is important to shift from the idea of ​​"being stolen" to the idea of ​​"finding and connecting differences." Differences are important for innovation. Friction doesn't occur when the same things are used.

I believe that the difference in positions of consultant, engineer, and scientist will create a basis for the evolution of technology, the update of values, and the update of the common foundation of the "common ground." The interesting thing about the concept of common ground is that it is not yet complete. When everyone puts out ideas, it creates a state where your ideas are part of the foundation.

This year's award invites people who can work together to create such a community, and the person who comes up with the best idea among them is awarded every year. I would like to invite as many people as possible to apply and create an experience where they can meet new people.

The Technology Laboratory is located at 1-1 Otemachi, Chiyoda-ku, Tokyo. It is said that it is also an attempt to expand the network from the "center" of Tokyo to the world.

──This year's award seeks to discover social issues under the theme of well-being and resilience and connect them to technology, or to reread one's own technology and connect it to social issues. Is innovation actually conceived from technology or issues?

Kanade: "Solving something that doesn't have a problem" already has a contradiction in terms of words, doesn't it? I think there are three basic conditions for innovation: idea, feasibility, and social issues. It is important that this triangle is well balanced. In today's world, it is difficult to think of innovation that is unrelated to the problem setting of what should be solved. I think that the keywords of this time, such as resilience and well-being, express a contemporary understanding of social issues.

Of course, there are many innovations that seem to have been born without social issues. However, it just happened that the researcher himself or beside him was someone who had a wonderful sensitivity to social issues, so it went well. You mentioned "replacement" earlier, but I think this is a very important keyword. The fact that when a technology is born, it is possible to reread the connection with social issues afterward is proof that the issues are driving innovation. This is because a new perspective has made people recognize the new value of technology.

──From that perspective, please tell us about any innovations that left an impression on you.

Kaneide Let's talk about when the world's most used algorithm called Quicksort was born. The Turing Award-winning computer science theorist who invented it, Antony Hoare, was originally involved in a project to machine translate Russian into English after World War II. To do this, we first need to look up the words in the original Russian text in a dictionary.

Dictionaries in those days were on magnetic tape, so to find a word beginning with Z, you had to advance the tape to the end. If you then draw a word starting with A, you have to rewind to the beginning. Therefore, if you look up all the words in the sentence in advance and arrange them in alphabetical order, the process will be faster. However, if there are about 10,000 words, it will take quite a while to arrange them in alphabetical order. From that challenge, Anthony invented a completely new sorting algorithm called Quicksort, which has been used in fields that require all sorts of data.

Michiaki Matsushima, editor-in-chief of the Japanese edition of WIRED, who also serves as a judge for the ``WIRED COMMON GROUND CHALLENGE'', acted as the interviewer for this interview. Clockwise from top left: Takeo Kanade, Michiaki Matsushima, and Shinichiro Mihari.

Takeo Kanade's ``Secret Places''

──Lastly, what social issues and technologies have you been interested in lately?

Sanji: I think it's interesting to see how the world after death is understood. Thinking about what will happen to the world after we die and how we can make people around us happy leads to well-being. For example, it should be possible to think that well-being is the flip side of the question, "What did you leave behind for the world?"

I think there will be services that leave records on the blockchain, revive humans from the afterlife, and reproduce them in VR. I can fix it. I think it would be interesting if such a slightly philosophical theme was submitted.

Kaneide: Based on my own experience working with artificial intelligence (AI), I think that most problems that humans consider difficult can be solved. Based on that, I would really like to see the theme of "Humans can't do it, but computers can do it for the first time." I think there are a lot of possibilities here.

Also, from a different point of view, I often think that what people have done traditionally has some meaning. Traditional culture, traditional arts, traditional industries... They still exist today because everyone has recognized the value of human life.

There is an AI research institute in the production area of ​​Kyoto's traditional industry "Nishijin textile". AI has both aspects of analysis technology and improvement technology.

Until now, civilization has been interpreted based on human intuition. I can't. This may be a little-known spot.

Period: January 2022 to February 2022 Final screening and award ceremony: April to May 2022

Eligibility: Age, nationality, and gender. Both adults and students can apply, and both individuals and teams can apply. Students from various backgrounds can apply, including university students, graduate students, researchers, engineers, startups, ventures, entrepreneurs, businessmen, architects, designers, creators, and programmers.

Submissions: Title of the challenge / Description of the challenge (approximately 400 characters) / Explanation on technology (approximately 200 characters) / Explanation on implementation (approximately 200 characters) / Explanation on globality (approximately 200 characters) / Contents of application Detailed description [Optional] / Project reference materials (images, document materials, videos, etc.) [Optional] / Profile (about 200 characters)

Period: 2022/1/12 (WED) ~ 2022/3/6 (SUN) Sponsored by: "WIRED" Japanese version Co-sponsored: Institute of Industrial Science, University of Tokyo Sponsored by: PwC Consulting LLC

[ Special site ]