The robotics ecosystem in Japan is amazing. And confusing.

It’s a collection of crazy ideas, odd creations, and true breakthroughs. And despite the combination of fawning prise and snide skepticism that Japanese robotics evoke in the international press, only time can really separate the true breakthroughs from the dead ends.

Today, we sit down with Tez Sawanobori, the founder of Connected Robotics, and we talk about how robots are being adopted in the restaurant industry here in Japan.

Connected Robotics already has two lines of consumer-visible robots being used in restaurants in Japan, and the reaction from the owners, the employees, and the customers has been overwhelmingly positive and quite a bit different than similar experiments run in America.

We talk about the strong economic and social pressures affecting the adoption of robots in restaurants and discuss the changes he had to make before chefs and robots can really work side by side.

It’s a great conversation, and I think you will really enjoy it.

Show Notes

  • The real reason we need robots chefs
  • The unlikely founding of Connected Robotics
  • Why the restaurant business is so hard to disrupt
  • Looking at the real economics of food prep robots
  • What’s holding back robotics in restaurants
  • Can robotics really solve the labor shortage in Japan?
  • How Japanese employment practices make it harder to use robots but increase the need for them
  • How Japan can catch up to the US and China in robotics research
  • The best way for American and Japanese robotics engineers to work together
  • The future of foreign workers in Japan

Links from the Founder

Leave a comment


Welcome to Disrupting Japan, straight talk from Japan’s most successful entrepreneurs.

I’m Tim Romero and thanks for joining me.

Today, we’re going to be talking about the OctoChef.

“What the hell is the OctoChef?” you might ask, and that would be a good and quite reasonable question.

The OctoChef was created by Connected Robotics and it’s a robot that makes Takoyaki, and we’re going to sit down with founder Tez Sawanobori and talk about why it’s important.

It’s important to understand that the OctoChef is not just some crazy side project of Tez and the team, although I guess it was the very first time I met them, but no, now, the OctoChef is being used in both small scale, single restaurant installations and industrial scale factory installations.

Tez and I talk in detail about how Japanese react to robotics and work with robots very differently than westerners do. We also sit down and eat some pretty good robot-cooked Takoyaki and take a hard look at the question of whether the OctoChef is just a novelty or a fad, or if on the other hand, it’s solving a real problem.

The answer turns out to be yes but the reason why is pretty surprising.

But you know, Tez tells that story much better than I can, so let’s get right to the interview.


Tim: So, I’m sitting here with Tez Sawanobori from Connected Robotics, the maker of the OctoChef, so thanks for sitting down with me.

Tez: Yeah, thank you for having me in this great show.

Tim: Thank you. What is the OctoChef?

Tez: OctoChef cooks Takoyaki. Takoyaki is octopus ball popular in Japanese festivals, you see a lot of Takoyaki stalls.

Tim: Yeah, the round little – and they’re awesome, it’s great food.

Tez: Yeah, yeah, it’s a popular Japanese fast food, and a robot cooks Takoyaki and it’s from pouring oil to serving to the dish, all the process the robots do.

Tim: Okay, so in the process, the humans still have to create the batter?

Tez: Yes.

Tim: And, I guess deliver the cooked Takoyaki to the customer.

Tez: Yeah.

Tim: The robot handles everything else? And you actually launched the first commercial use of this in Nagasaki a while ago, so how’s been the reaction of the customers to the robot Takoyaki chef?

Tez: Customers look very pleased to eat Takoyaki cooked by robots and this is a special occasion for them, and especially in Nagasaki, there’s a theme park called Huis Ten Bosch, so they enjoy looking at robots working.

Tim: Okay. Yeah, I want to talk about like the business model in detail in a few minutes, but before that, let’s back up a bit and talk about you.

Tez: Okay.

Tim: So, you founded Connected Robotics in 2014 but before that, you started a restaurant right out of college, right?

Tez: Yeah, right. I studied Robotics and Computer Science in the University of Tokyo and later, I went to Kyoto University and researched Computer Science, but when I graduated, I wanted to start my own business. Of course, I tried some web services, but I didn’t get satisfied with this kind of IT.

Tim: Really?

Tez: So, first thing I wanted to do was making a cool restaurant.

Tim: It’s a lot of work! What kind of restaurant was it?

Tez: Oh, at first, I really liked jazz music so I wanted to make like a jazz club, but at the same time, I want to make some family-friendly restaurant with good music. I really want family and small children to come to our restaurant and enjoy our food and atmosphere.

Tim: So, what happened to the restaurant?

Tez: Oh, actually, I didn’t have much money, obviously, because .. –

Tim: You just graduated, right

Tez: Yeah, yeah, I just, yes, started working after grad school, so I planned several restaurants and brought my idea to many presidents of restaurant chains, and one of them was very interested in me and the plan, so I joined this company and set up several restaurants, including Italian, American, Japanese food restaurants.

Tim: Oh, wow.

Tez: Yeah.

Tim: Why did you decide to leave the restaurant business?

Tez: Until then, I had several experience in restaurants, like during summertime, I worked in izakaya and my grandparents owned their restaurant, and I thought I know a lot of restaurants, but actually, I worked in the restaurant and found that, well, all the work are very hard, very long.

Tim: It’s hard, hard work.

Tez: Yeah, yeah, so I found myself working like, 100 hours per week.

Tim: Oh, wow.

Tez: And exhausted. So, I wanted to come back to technology side. I belonged to a robotics club in the University of Tokyo and won a competition called RobotCon for the first time for the university.

Tim: How many years were you working with the restaurants?

Tez: Oh, just one year.

Tim: Oh, just one year?

Tez: Exactly one year, yes.

Tim: Okay, and so during that time, you’re still active and interested in robotics?

Tez: Oh, yeah, yes. Yes, for sure.

Tim: So, did the idea for food robots come to you while you were working at the restaurant and that’s why you quit or you just had to get out of the restaurant?

Tez: Oh, yes, yes. Many people, while I was working in the restaurants, recommended me to start this kind of business, robot restaurant, but at that time, I thought it was too difficult to deploy robots in restaurants because all the operations are so complicated. It’s really difficult.

Tim: Yeah, yeah, cooking is an incredibly complicated process and kitchens are tiny.

Tez: Yeah, usually, tasks are not so organized and oh, yeah, so I couldn’t find any good work for robots. Just, I felt overwhelmed.

Tim: But you gave it a try anyway.

Tez: Yeah, yeah, yeah.

Tim: Actually, I saw you guys at Maker Faire a couple years ago.

Tez: Oh, yeah, yeah, yeah, yeah.

Tim: And it was the same – I mean, it has gotten a lot better since then, but it was the same Takoyaki-making machine.

Tez: Yeah, yeah, yeah. At that time, we showed our first prototype.

Tim: Maker Faire is actually, it’s a really fun place to go. There’s so much creativity in robotics and automation in Japan right now.

Tez: Yeah, yeah, that’s true. Yeah, everybody is trying and challenging something new. Some of them look very strange, but –

Tim: Yeah, there’s a lot of strange crazy stuff, but to be fair, when I first saw you guys there, I was thinking, okay, Takoyaki-making robot, that’s kind of strange, you made a business of it.

Tez: Yeah, at first, my ex-colleagues always told me that, “It’s really strange,” and “You shouldn’t do that,” and “It’s crazy!” and something like that.

Tim: Yeah, and I’m sure you still hear that occasionally, but let’s talk about the business model. So, what is the revenue model? Are you selling robots or selling services, or selling Takoyaki?

Tez: Okay, basically, we don’t make the robot itself. We integrate this robot and make it as a package and sell this to our customers, that is restaurants. Firstly, we sell this package, but from the second time, we will lend this robot and charge monthly.

Tim: Okay, and what does the OctoChef cost monthly?

Tez: ¥250,000 per month.

Tim: Okay, so $2000 a month?

Tez: Yeah, roughly.

Tim: Okay, I know the OctoChef makes Takoyaki, but does it do the work of like, a full-time staff? $2000 a month is not much compared to what you have to pay employees and things to do same work, but how much work does it actually do?

Tez: Actually, Takoyaki requires high skill, and Takoyaki restaurants usually train the employee, and it costs a lot. It takes time and costs a lot. With the robots, they can save this training and hiring process.

Tim: For the restaurants that are buying this, is the motivation saving money or is the motivation the novelty, the kind of the cool factor of people watching the robot make Takoyaki?

Tez: Yeah, of course, at first, their first purpose is to get novelty and show this robot, but actually, restaurants want to have more reliable labor. Yeah, yeah.

Tim: Yeah, yeah, I mean, Japan really is in a bit of a labor crisis right now. It’s hard to fill up positions.

Okay, so if the restaurants are paying Connected Robotics $2000 a month for an OctoChef, how much money are they saving in labor costs?

Tez: First release of OctoChef in Nagasaki, the restaurant is run by one operator. Usually, they have to hire like, three or four, even more than five people for this Takoyaki, and so they can just put in this position one person. Yeah, they can save one or two other employees.

Tim: Okay, so that’s a very strong proposition. It’s not just the novelty of having the robot, it actually makes sense. Do you think we are going to be seeing more robot chefs? Because you mentioned that like, we were talking before, cooking is incredibly complex, and Takoyaki, I mean, it requires skill, but it’s a pretty well-defined set of motions, right? So, do you think we are going to be seeing more and more robot chefs for other kinds of food?

Tez: Yeah, I think it’s a matter of time, but it takes a long time depending on how big they are cooking or the skill is, of course, cutting sashimi, it’s very difficult, but we highly focus on specialty restaurants like Takoyaki, Yakitori, or something like that, like Okonomiyaki. In Japan, we have a lot of family restaurants who served various cuisine, but we focus on single cuisine like Takoyaki.

Tim: Right, right. I mean, automation is taking over everything and that’s good, but food somehow seems different. I mean, sure, sure, industrial food preparation is completely automated, but somehow, restaurants, there’s just something very human about making food and I don’t know, I kind of like watching the chef flip the Takoyaki or fan over the Yakitori, even if it doesn’t necessarily make it taste any better. Do you know what I mean? It’s like, part of that experience.

Tez: That’s a cultural issue. I heard that in the United States, people feel safer that they found sushi robots in sushi restaurants.

Tim: Really?

Tez: Yeah, I heard so, but in Japan, if customers find this sushi robot, they think this is a cheap one and it’s not reliable and won’t taste good. So, this is a cultural gap. Yeah, of course, it’s very important how they feel about taste. Of course, everything affects the taste, and so we really care about that. We don’t say, “This is automated cooking,” or something like that. We always insist that this robot helps workers, so that the workers can get more creativity, easier working place.

Tim: So, as we see more and more of these cooking robots, do you think they are going to be out in front of the customers as novelties or sort of hidden in the kitchen, just operating behind the scenes?

Tez: Actually, we have three products, and one is like, Takoyaki, ice cream, Yakitori, and cheesecakes, something like that. This is for front end robot. Yeah, robot in front of the customers and cooks just in front of the customers. This is one thing. Second one is dishwasher. This one is put inside the kitchen and customers cannot look at these robots working, and sometimes, especially for convenience store and cooking in the counter space, honestly, I’m not sure which one is better. Yeah, but we are sure that people somehow want this novelty. Of course, at first, but for restaurants owner, they want both novelty and making kitchen have less trouble and complexity.

Tim: Do you think that these kinds of robots, not just Connected Robotics robots, but these kind of robots, do you think will actually contribute to kind of solving the labor shortage in Japan?

Tez: Yeah, I believe so, yeah, yeah because now it’s been two years since we started this cooking robot project, and we already have got many requests from all over Japan and from all top 10 restaurant chains in Japan, so they really need this robotic system. So, I’m sure it will solve the labor shortage in Japan.

Tim: Yeah, it’s fascinating because everyone always talks about the shrinking population as pushing down GMT, which it does, but the other side of that is increasing productivity by using things like robots. So, what are the barriers? What’s preventing robots like OctoChef and similar robotics from being even more widespread in Japan?

Tez: One is organized operations in the kitchen or even at a restaurant. Usually, restaurants have really complex, and especially in Japan, one worker has a lot of tasks. Maybe in the United States, I think they put a person to a position, and this person has distinctive obvious tasks, but in Japan its more ambiguous and people have to do everything, even whole staff have to work in the kitchen. Kitchen staff have to go out and serve dishes to customers on the table.

Tim: Especially at smaller restaurants.

Tez: Yeah, yeah, yeah. So, first thing to do is to organize these tasks. Yeah, and second thing is, of course, cost effectiveness and efficiency.

Tim: So, we talked a bit about how the customers were reacting to the robots. What’s been the reaction of the staff, the other workers?

Tez: That’s the first thing we really care about because we want to lessen the labor. Actually, we’re always watching how they behave with the robot. We are really busy in taking care of staff to cooperate with the robot because for them, they’re not technicians. It’s really a funny thing, is working beside the robots. First few months, every day, we take care of the staff.

Tim: So, there’s a lot of training?

Tez: Yeah, training. Not just training but watching from video and giving advice, and like, everything.

Tim: But overall, the staff is really willing and interested in working with OctoChef?

Tez: Yeah, yeah, obviously.

Tim: One of these things that’s so different in the US and Japan, so I think it was last year, the Flippy robot? So, Flippy, it was scheduled to be rolled out in like, 52 Cali Burger restaurants, and they ran one shop for one day and shut it down because the staff just – mutiny, they rebelled, they didn’t want to work with the robot, and Japanese staff seem to have no problem with it, so why the big difference?

Tez: Yeah, in Japan, people think robots are friends, and I think, I guess in the United States, robots are just tools or it could be an enemy, like Terminator?

Tim: Yeah, in the US, robots are the Terminator. In Japan, it’s Doraemon.

Tez: Yeah, yeah, yeah. In Japan, yeah, like Doraemon, Astro Boy? Yeah, they are friends, but in the States, it’s a potential enemy or a potentially dangerous existence. So, I think in the United States, people think this robot will take away my job, but in Japan, this robot is helping me, so this is my friend.

Tim: That is interesting, and you see that attitude towards automation across all industries in Japan and the US. This is such a different approach. Actually, Connected Robotics, you have a really international staff. Your team is really international, and Japan kind of has this reputation of being really strong on the robotics hardware side, and overseas engineers kind of have the reputation of being really strong on the software side. Is that kind of how it has worked out for your team?

Tez: Oh, yeah, that’s true, yeah. Actually, I put an advert to hire for engineers, and I got more than 200 applications from all over the world. I actually hired two guys from Europe and they said that in Europe, they don’t have such place to use their knowledge learned in university. I mean, robotics knowledge, so they want to make full use of robotics, but they don’t have suitable company in the country, so they look for robotics companies in Japan. So, I think there lies greater opportunity for Japanese robotics companies to hire foreign roboticists.

Tim: Yeah, it’s really interesting that Japan, for 30 years, Japan was leading the world in robotics, and over the last 10 or 15 years, we have seen a lot of American companies and Chinese companies really running out and taking the lead now that software is so much more important, but it seems like just in the last four or five years, Japan is putting a lot of effort into catching up; there’s a lot of really interesting robotics innovation going on now.

Tez: Yeah, yeah, that’s true, yeah, that’s true, but Japanese robotics engineers are more interested in hardware. That means creating their own robots whereas engineers in the United States or in the US like to implement software AI program. So, it’s a good comparison. Japanese robot engineers are really interested in making robots while foreign robot engineers maybe want to implement AI into robots.

Tim: Yeah, the old kind of stereotype still is true.

Well, listen, Tez, before we wrap up, I want to ask you what I call my “Magic Wand” question, and that is, if I gave you a magic wand and I told you that you could change one thing about Japan – anything at all, the education system, the legal system, the way people think about risk, anything at all to make it better for startups and innovation in Japan, what would you change?

Tez: That’s a really interesting question. With this magic wand, I want to make a city where people from all over the world can really come and live, in diverse cultures, because to make a new culture, we have to mix many elements, many types of people, and let them do whatever they like, really.

Tim: The idea would be just to get more foreigners bringing different opinions of different ways of thinking, that is interesting because one of the things I’ve noticed is a lot of regional innovation centers in Japan are trying to get foreign entrepreneurs and foreign students, not just Japanese.

Tez: Yeah, at the same time, I really want to conserve Japanese pure culture, so I want both, like we have Tokyo, but at the same time, we have Kyoto. I lived in Tokyo and Kyoto, and they are very different atmospheres, culture, and both experiences are really foreign for me. So, my opinion, I think Tokyo should get more diversity,

Tim: But within your own company, for example, has that diversity been really beneficial?

Tez: Yeah, yeah.

Tim: But I’ve noticed – so, Connected Robotics and a lot of Japanese startups have mixed teams, both Japanese and foreign members, but it’s interesting, you mentioned about the importance of tradition because most of the time, I mean, almost all the startups in Japan are trying to develop a very kind of westernized startup culture, not a traditional Japanese business culture at all. So, how do you take that foreign influence and foreign creativity, and different ways of seeing things in different ways of thinking, but also protect that tradition that you mentioned?

Tez: Yeah, in my company, we have Japanese engineers and foreign engineers, and I’m really interested in looking at how they work. Japanese engineers try to do everything and foreign engineers try to focus what he or she is supposed to do. There is a different style of working, but both are important, and with this combination, we can solve any problem.

Tim: Yeah, and it seems to be happening. The government has made the visa laws much, much looser, it’s much easier to get a visa to come and live in Japan, I mean, before, as a start up, you couldn’t even sponsor a visa. Now, it’s really simple.

Tez: Yeah, that’s true. Yeah, yeah. The government is inviting high-skilled talent from all over the world, but I think at the same time, they should accept more various types of people, not just evaluating their skill, like –

Tim: So, what other kind of people or what are the kind of views that you think that Japan should be bringing in?

Tez: They should think more about diversity, so not just one skill or one specialty. I think we should accept many types of people.

Tim: Well, hopefully, we will see that happen because it’s becoming a very welcoming country to foreigners.

Tez: Yeah, I think one thing that in Japanese restaurants, food are so cheap. I think it’s one of the big reasons. Yeah, but it’s becoming more and more difficult for restaurants to keep this price.

Tim: And, convenient stores a lot now, a lot of the employees are foreign.

Tez: Yeah, yeah, and tasks are so complex. One worker has to buy items and manage items and do cashier, and cook fried chickens, and everything.

Tim: Hey, listen, Tez, thanks for sitting down with me, I really appreciate it.

Tez: Yeah, thank you very much.

Tim: And, we’re back.


Cooking is a uniquely human activity and not just for the reasons that Tez and I talked about, but I’ll get back to those in just a minute.

I mean, cooking is something that in general is exceptionally difficult for artificial intelligence to do. Take for example, slicing a tomato. It’s not particularly hard for us and a child can do it as soon as they are old enough to be trusted with a knife, but for a computer, slicing a tomato is hard. No two tomatoes are exactly the same shape or size, and the way you cut around the stem is different for each one. Some of the tomatoes will be bruised or damaged, and you will need to cut around those defects.

Now, you and I, and even our knife-wielding toddler can do this without much conscious thought at all. We can debate whether this actually demonstrates that AI is particularly bad at these kinds of loose generalization tasks or whether it demonstrates that we humans are particularly good at them. I’ve heard arguments on both sides, but I don’t think it really matters.

The end result is that AI is just starting to make inroads into cooking, and where it is making those inroads is quite telling. You see, unlike our tomatoes, Takoyaki is very standardized. The size, shape, texture, and heating is very much the same from one little ball to the next.

In a similar vein, after the interview, Tez was showing me around another kind of robot that Connected Robotics is making. This one serves softserve ice cream, a fairly standardized or standardizeable process, but AI will get better in time.

I have no doubt that it will eventually learn to competently handle a tomato, but the bigger issue I see here, and it’s one that Tez and I disagree about, I think there is something special, and again uniquely human about cooking and eating a meal. It is a shared experience with not only the people you eat with, but the people who prepare it for you.

But maybe I’m wrong. Maybe this is a generational difference between me and Tez, or maybe it’s a cultural difference between Japan and America. We will all find out in the coming years. But to me, a shared meal is not just about the people you eat with. Part of the experience, part of the value and part of the connection we feel to our food is the people who prepare the meal for us.

It’s something that is deeply human. It’s something primal.

If you want to talk more about robot chefs and the future of the human experience that is the shared meal, Tez and I would love to talk with you. So, come by and let’s talk. If you leave a comment, I guarantee that Tez or I, or maybe even both of us will respond.

And hey, as most of you listeners know, Disrupting Japan is completely free. We don’t do crowdfunding and we don’t have a Patreon site or anything like that. I do it simply because I love making the show and a lot of fans love the show and they have asked me how they can help support it, and I appreciate them. I mean, I love that, and if you want to support the show, the absolute best thing you can do is to tell people about it. Maybe share a post on social media or just tell a friend about Disrupting Japan in a plain old-fashioned conversation sometime. In our world of click bait headlines and fake-it-till-you-make-it founders, you would be amazed at how powerful an honest recommendation can be.

Most of all, thanks for listening and thank you for letting people interested in Japanese startups know about the show.

I’m Tim Romero and thanks for listening to Disrupting Japan.