img/workplace/masami_yuki/6/1.png img/workplace/masami_yuki/6/7.png
img/workplace/masami_yuki/6/1.png img/workplace/masami_yuki/6/7.png

【Hirohisa Inamoto】
IVI Development Department
Core Technology Development Center
Research & Development Group

Born 1975 in Osaka. Joined Ricoh in 2002 upon finishing his graduate program at Osaka University. Since then, has been primarily engaged in image processing technology R&D. Led projects to develop technology to improve copier image quality, and to develop new features for scanners using image recognition. Currently leading R&D efforts to create image-quality prediction technology for the inkjet printers used in improving in-house design work.

【Masami Yuki】
Manga Artist

Born 1957 in Hokkaidō. Debuted in 1980 in Gekkan OUT (published by Minori Shobō) with the series The Rival, while holding down a corporate day job. Quit his company job after six years in 1982 to work as one of the principal artists on the Weekly Shōnen Sunday magazine. Won the 19th annual Seiun Award for best comic of 1988 for Kyūkyoku Chōjin R. Won the 36th Shogakukan Manga Award in 1990 for Mobile Police Patlabor. He has continued to work on long-running series such as Jaja Uma Grooming Up! and Birdy the Mighty. Yuki has moved a lot of his work over to digital of late. Yuki is also active on Twitter (@masyuuki) and has gained some following there. His new series, White Curtain Chronicles, is now running in the Big Comic Spirits magazine.

Wow, 15 years since Patlabor!?

Inamoto: But robotic engineering hasn't progressed as far as shown in Patlabor. And there might not be any vehicles like that in 2036 that people could easily ride in. This is definitely a field where I'm looking forward to the next generation [of development].
Yuki: Yes, exactly, looking at the folks building robots now is really interesting. The first people to build something robot-like were the generation that wanted to make Mobile Suit Gundam. So now it looks like the young people who read and were influenced by Patlabor are starting to have a go.
Inamoto: When the series started in 1988, the scene was set in 1998, wasn't it? And now reality has moved another 15 years past that.
Yuki: Did you know that Osamu Tezuka's Astro Boy's birthday was April 7, 2003? That's already more than ten years ago, yet Astro Boy is beyond what current technology can do. But then, on the flip side, some manga episodes showed scenes [of this sci-fi future] where mountain villages hadn't yet been connected to the power grid. Nowadays, that kind of hodge-podge, hit-or-miss feeling is really interesting.

The roles of the manga artist and the researcher

Inamoto: Ricoh's business has changed a lot since its founding in 1936. That's nearly 80 years of history, and in looking ahead to the company's centennial in 2036, I think we need to be able to think even more outside the box, so to speak. Along those lines, are there any technologies you hope to see become a reality by 2036?
Yuki: Hmm, when I'm asked that, I can't bring anything to mind. There's a part of me where I don't really know what I want until I've drawn it first in a manga.
Inamoto: Really?
Yuki: Japan's sci-fi manga are often described in the wider public as really insightful, but as time passes, there are lots of areas where the real world doesn't quite catch up. Put another way, we should be writing more attractive manga for the things we have high hopes and dreams about. If we do that, maybe then the odd researcher will help make that dream a reality. [laughs]
Inamoto: We need to keep our antennae switched on to pick up that message. After all, it's our role [as researchers] to make those dreams a reality. This game of tag between creative types and researchers is a lot of fun. [laughs]

* All names of companies, products, and services used on this website are the registered trademarks of their respective companies.


Hitting on what's "just about right" is the hardest

Yuki: But then, a non-expert like me doesn't really understand what image recognition is. For example, how does a computer recognize human faces in an image?
Inamoto: In the research I've been involved in, we make a statistical judgment based on the brightness distribution of an image, and how image elements are aligned. Other research angles use color, such as by recognizing an area with a lot of skin tone as probably being a face. Things that are easy for humans are still quite difficult for computers to do.
Yuki: I've heard that the biggest difficulty for computers is deciding when something is "just about right." But even then, might computers outstrip humans in these areas too, some day?
Inamoto: I have my doubts. The human brain is no match for a computer when it comes to straightforward calculations, but we are the clear winners when it comes to recognition processing. We make use of this capability every day without even being aware of it, but it's really a highly advanced ability. Computers have been getting really close, but I think it will take some time yet before they can completely overtake us.

Moving too quickly presents other problems

Yuki: Okay then, so I guess we have to wait for some kind of breakthrough. Maybe artificial intelligence isn't quite there yet to the point of being really impressive.
Inamoto: That said, there have been examples of implementing a large-scale reproduction of old technology that modeled the brain, using only modern computers, and getting dramatically good performance. Better computer performance is really important, and we might not even need a breakthrough in recognition technology. But in terms of impressive, what was the most surprising technological development that you've encountered in your life?
Yuki: That would probably be the mobile telephone. When we were doing the Mobile Police Patlabor series, I had no idea that cell phones would proliferate the way they have.
Inamoto: So wait, you're saying that your manga didn't show any cell phones? That surprises me somehow.
Yuki: Yeah, cell phones didn't take off until after the series finished. That was a good thing. If that had happened while we were doing the series, well, the real world advancing that quickly... that would have been a bit of a problem [laughs]. But then again, we do have cell phones now, so we can't just skive off work and go to the movies anymore.

Daring to degrade

Yuki: The copiers in our office have actually always been Ricoh machines. The printing quality has always been very crisp and clear, and we really like them.
Inamoto: That's an honor! But I've heard that even manga are increasingly being produced digitally. Aren't there fewer opportunities to actually print out manga drafts on paper?
Yuki: No, even when we work digitally, we do ultimately want to put it on paper to check the results. Readers get our final work as a printed product, after all. And then, it's a bit odd, but it's easier for us to spot typos on paper.
Inamoto: Yes, I've read research data saying the same thing. When comparing handwritten versus printed writing, it seems that handwriting is easier to visually parse.
Yuki: Oh right, you've been working on research into image recognition so of course you'd know all about this.
Inamoto: No, no, but given my background, your idea for the future of degrading the resolution of a holographic projection depending on the priority was a real epiphany for me.

"This isn't real, you know."

Yuki: When I'm drawing, I really can't stand the loneliness when working on my own, so I think it would be great to be able to share even a virtual space with an assistant who might even be working remotely, but there really wouldn't be any need to see each other in high resolution when we're working.
Inamoto: Yeah, so you're saying it'd be good to feel like they're really there.
Yuki: Yes. Say, like, folks who are comfortable around each other, they might not get spruced up for work, and it'd be good to be able to hide that, too. [laughs]
Inamoto: Among us researchers, when it comes to images, we've been working on how high-quality and how high-speed we can make things. So being able to shift the emphasis in images as you propose strikes me as a really fresh idea.
Yuki: Maybe that's just the kind of idea a manga artist would have.
Inamoto: Even so, this would be a system that would recognize both virtual images and real-world images all together. That's precisely why we would need to clarify that "this isn't real, you know."

In this our second session, we've brought together Masami Yuki and Ricoh engineer Hirohisa Inamoto for a discussion at Ricoh headquarters in the Ginza district of Tokyo. Yuki further describes how people will work in 2036, based on his vision of the future. The conversation is broad-ranging and dives deep, touching on subjects from Yuki's Mobile Police Patlabor manga to computer recognition processing. Just what kind of imaginations were sparked here?

Masami Yuki
Manga Artist
Hirohisa Inamoto
IVI Development Department
Core Technology Development Center
Research & Development Group
img/workplace/masami_yuki/4/1.jpg img/workplace/masami_yuki/4/1_2.jpg img/workplace/masami_yuki/4/1_3.jpg

Work styles in 2036 – virtual commuting?

In any future, people are going to want to engage using their real senses.
Current technology looks like it has evolved by extending familiar input methods,
like using your voice to control your smartphone, or entering text using handwriting,
so I really don’t think that the traditional commuting work style will have disappeared by 2036.
That said, even folks working from home or other remote locations will be able to collaborate with their colleagues
just as effectively as face-to-face, be it in real or virtual spaces…
I think that might be the kind of work style that we see develop.

* The text on this page has been composed by Ricoh based on interviews with Masami Yuki.

Advanced AR glasses project
remote colleagues in 3D
Her presence is really like a wraith, and she's not actually there. However, as illustrated here, looking through the advanced AR glasses shows the wearer a window onto a remote scene. You'd be wearing the same glasses that she is. These would show you a holographic life-size projection of people in remote locations.
Change the wallpaper seen
through the AR glasses
So again, this illustration shows the scene when looking through the advanced AR glasses. The temple in the background is also not really there. This is projection mapping, and the glasses are projecting the image onto the walls of the room. I really think that people's working efficiency would improve if they could change their wallpaper to match their mood for any given day.
Blur out other conversations?
Change display depending
on priorities
The man on the left is physically present, while the woman to his right that he's talking to is a remote colleague displayed by the glasses. People who aren't important to you right now could be displayed in a lower resolution to save energy. That's because the important thing is feeling their presence.
Synchronizing the work of
all the team members on
in-desk monitors
This desk is synchronized with remote colleagues' desks, so that all members of the team can see all of the work in real time. So this, together with the 3D projection from the AR glasses, helps to strengthen the sense that the whole team is sharing the same space.