Episode Transcript
Speaker 0 00:00:02 Hi, my name is Tim Beretta. Welcome to the robot industry podcast. We're glad you're here and thanks for subscribing. My guest today is Dave Johnson from DXI robotics out of Boston. I'm excited to have you, Dave, because I love talking about food and robotics and automation. So welcome to the podcast. Hello.
Speaker 1 00:00:16 Uh, welcome. Glad to be chatting with you, Jim. Uh, my name is Dave Johnson and I'm the CEO of DXI robotics. And we built Alfred, which is a robot sous chef, a robot, which is designed to go into commercial kitchens and assemble meals exactly as, uh, they would be made if you or I were making them in our own kitchen.
Speaker 0 00:00:43 Thank you, Dave. And this is Dave Johnson. We really excited about today's episode because if you know me and I don't really know you at all yet, I'm a big foodie and Dave, I think we have to call a will Healy the third for introducing us.
Speaker 1 00:01:01 Yeah, it's always great to chat to somebody who knows a will Wilson, a great friend. We go back many years and he's a huge enthusiast for the industry.
Speaker 0 00:01:11 Dave, how did you get involved in robotics?
Speaker 1 00:01:15 Wow. So I've been building robots now for, uh, nearly 15 years. I started with unmanned helicopters, uh, working on a, a project which was building the longest endurance, uh, unmanned helicopter at the time back in the early two thousands. And, uh, from there transitioned into making atomic clocks, which is, I know it's probably not the usual way that people get into robotics, but there's a lot of, uh, uh, similarities between, uh, unmanned helicopters and, uh, it really crazy precision instrument like an atomic clock. And, uh, I also have a huge hobby and passion for food. So, uh, as I was, uh, uh, seeing a lot of really interesting, um, uh, technology getting developed in the, uh, learning space, especially that, uh, we could recognize objects much, much better than, uh, than we used to be able to before I decided to, uh, quit the atomic clock business and start a, uh, robot chef company. And I've been, uh, working on this now for oh, nearly five years in duck side has been around for three and a half as an independent company.
Speaker 0 00:02:31 And that's very exciting. So tell our audience about more decks I, and about Alfred.
Speaker 1 00:02:35 Yeah. So we're, uh, at tech side, we're building a robot sous chef, which is, uh, a robot assistant that's, uh, really what scifi has been promising us for years about robots, which are going to, uh, make you food and cook you dinner. Uh, it's designed as a, an assistant to humans in really any kitchen. Uh, our first market is commercial kitchens. Uh, what the product, uh, is, uh, at its core fundamental is it's a, a robot arm which uses utensils. So think, uh, ladles spoons, salad, tongs, uh, scoopers dishers, uh, spatula is really anything that a human chef would use to manipulate food exactly in the same way that you would do it if, uh, you were a chef, uh, making a salvage for instance. So, uh, what this looks like is we put a robot arm in the kitchen, uh, mounted to a refrigeration unit or to a grill or a stove, or really any workstation that you would have in the kitchen. And then the robot, uh, using, uh, computer vision, uh, locates itself in the environment recognizes the foodstuffs and, uh, for the recipe that is going to be making and then, uh, goes through and assembles a meal, uh, or cooks a steak or, uh, operates a fryer, uh, exactly according to the, uh, uh, the food safety guidelines and, uh, in the recipe that the chef gave him.
Speaker 0 00:04:11 And one of the big motivators for you of course, to do this is there's a huge amount of people that coming back to work or even employed in the food service business.
Speaker 1 00:04:22 Yeah, that's, uh, it's tremendous of right now, the situation is, uh, perhaps even more serious than everybody, uh, realize it's, uh, there's just an incredible shortage in folks that, uh, uh, are working in kitchens. And, uh, we just, we want to build something that would, uh, basically enable the, uh, people to do their jobs better and faster and prepare, uh, more meals than they would be able to do by themselves. And it's, I would say that this is, uh, uh, COVID is only, uh, uh, accelerated that trend, which was existing before,
Speaker 0 00:05:06 And you're using a collaborative robot and this, for some of the people listening to podcasts, this is a robot that won't hurt you.
Speaker 1 00:05:15 Correct. And we've even gone above and beyond that. So one of the really cool things about, um, I would say our, our general approach to this problem is that we, uh, have built a robot, which is really aware of its surroundings in a way, which is, uh, uh, different than, than what you have in even, uh, typical, uh, cobalt installations. So the Alfred is really designed to sit in an environment, which is not really very well-controlled, uh, kitchens are noisy, messy, uh, chaotic spaces, and do never really, uh, can expect to the same precision that you might have on an assembly line. Therefore, we have to always be observing our environment and because of, uh, breakthroughs really, uh, in computer vision and machine learning to recognize objects, we actually know what's around us, uh, really quite well. So we know when, when, uh, people are in the workspace and we know where the objects are in the workspace. And we're also, as you say, we're a cobalt. Uh, so we're continually sensing for collision. So we have this layered, uh, approach where the robot is actually actively avoiding, uh, contact. So it actually actively avoids you, if you, you, you get this way, but then if you go out of your way to try and pit the robot, uh, it's smart. And then it, it, uh, uh, senses the collision and recovers and plans away from the collision.
Speaker 0 00:06:49 It's so very exciting. What are some of your big challenges that you've had and that you maybe dealing with now?
Speaker 1 00:06:56 Oh, yeah. As well as I alluded to, uh, there's a lot of really cool things that we do because we're in the kitchen, but, uh, it's also, uh, an enormous challenge to build a robot, which is that flexible and that dynamic. And so when you add a lot of this really cool cutting edge, uh, technologies such as, you know, a real-time path planning and obstacle avoidance and things like that, uh, it creates a very, uh, uh, complicated system. And, uh, the most important thing in, and probably the biggest challenge is really consistency because that's what, that's what restaurant owners really care about and why they want a robot in the first place is that you can add more consistency to the process. And, uh, so that's, uh, putting that complex system in a messy environment and then making it work day in, day out with no failures is, uh, uh, really the, the big challenge.
Speaker 0 00:07:57 It's absolutely fascinating. So why would I want a robot in my kitchen as a restaurant?
Speaker 1 00:08:03 Well, so you already alluded to the fact that it's very challenging to, uh, to find folks right now, but, uh, even beyond that, uh, you really want to, uh, make sure that the meals that your guests are, are receiving are the same, that they are always expecting, right? So you want a, a, uh, a consistent, uh, product, which has very little waste, so minimizing food waste, uh, and, uh, you want to be able to feed more, uh, people than you could do already. So this idea that the robot can assist, uh, your existing team and make them more productive, and then finally, uh, figuring out how to reduce costs and that, that comes from, uh, maybe offering service at odd hours when you would normally be open, because you can run the robot. Uh, it comes from, uh, like I said, minimizing food waste. So you're not, um, wasting food and thus reduce or improving your bottom line there.
Speaker 0 00:09:10 Nah, I think it's very cool. Now let's talk a little bit about the technology to control systems. You mentioned that you vision inspection.
Speaker 1 00:09:20 Oh yeah. So we do, we do everything based on vision, uh, which I think is, uh, uh, you know, an emerging trend these days that, uh, that you don't have to, um, know quite so much about the workspaces you might have had two previously. And so our, uh, our product is really built on, uh, this benchmark technology, which lets us manipulate materials, which, uh, deformed or change as you interact with them. So you can imagine that if you're, uh, scooping ice cream, as you're, uh, pushing this food into the ice cream, the ice cream is changing and moving away from this food and doing all sides of all sorts of crazy things, uh, which makes it, uh, pretty hard to just use a, uh, a, uh, sort of a conventional a to B path in order to do that. So our, uh, our approach really had the robot learn by doing the task over and over again, and seeing the results, how to manipulate these materials.
Speaker 1 00:10:22 So we have actually a, a fleet of robots, which are, are running, uh, every day practicing and how to manipulate the materials. And, uh, every time they make a mistake, they learn from it and all the robots are learning from each other. And then, uh, we repeat this over and over again, you know, thousands and thousands of times. And then eventually the robot, just like a child has learned how to manipulate, uh, a different material. And it's also learned how to recognize that material. And so then you put that all together and now, uh, the robot is learning these skills and, uh, and then those are, uh, are tasks that, uh, the customer can use.
Speaker 0 00:11:03 So if I'm a restaurant owner and I've got an ice cream robot, which I think is very cool, and I think people would come to my restaurant because of the robot. Um, does that collective shared between all the other ice cream robots in your fleets?
Speaker 1 00:11:17 Yeah. So that's one of the big benefits is that the customers, if they want an ingredient, which another customer has already, uh, uh, requested, then they can just have it. So it's this network effect that lets you be as creative as you want. Uh, so if you're a chef and you say, oh man, I have this great idea for a, uh, a new ice cream sundae for instance, but boy, it needs, uh, bananas. And then it, it turns out that, uh, you know, another customer may have wanted, um, uh, bananas for a fruit salad. And because of that, uh, that sharing, uh, you can immediately get access to, uh, to the different ingredients in that, uh, allows us to, uh, basically have a huge library of, um, uh, other ingredients and build them up very quickly.
Speaker 0 00:12:12 So that's a good segue into my next question. And it's about data. How important is data then to this project?
Speaker 1 00:12:19 Oh, Dave data is everything. So as I, uh, I guess described earlier, right, the robots practicing all that data is crucial to how they, uh, to how the controller actually knows how to manipulate, uh, the ingredients. And then, uh, when we add new ingredients, right, we have to, again, repeat the process. And so every single time the robot does something, it's getting a little better. And by repeating that over hundreds and hundreds of thousands of times, uh, the robots actually get very good at performing their task.
Speaker 0 00:12:57 And so let's talk about robots as a service because this is your model, correct? Yeah.
Speaker 1 00:13:02 And we're, uh, we've definitely been, uh, piloting robots as a service with some of our restaurant customers because, uh, if folks don't know, right, restaurants are traditionally a very challenging economic, uh, business to get. Right, right. You can have really feast and famine kind of cycles. And, you know, most restaurant owners don't have a large sum of money lying around to purchase a cutting edge robot. But what they, what they are really good at is, um, uh, variable costs. So, uh, because if they can treat the robot costs exactly like their food cost. So if they sell more meals, they have to buy more food. If they eat more meals, they have to, uh, buy more time for the robot. And so it's a very natural business model for, uh, for restaurants, which is why, uh, why we chose it.
Speaker 0 00:13:58 And so let's talk a little bit about the low hanging fruit for a robot chef. What I'm thinking, I'm thinking about salads and burgers and pizza and ice cream, what else is, we'll hang it for?
Speaker 1 00:14:09 Well, uh, anything that can be made with any utensil is the way we, like, we like to think about the world in terms of utensils, uh, because that's what, uh, the robot, uh, you know, our whole thesis is that we use utensils to manipulate the world because that's the way human chefs do it. And so for, uh, for your examples, right, salads are great, right? You assemble them with, uh, with spoons and tongs and, um, uh, squeeze bottles for salad, dressing, uh, burgers you make with, uh, you know, a spatula, um, or also for tongs to select, uh, different ingredients. It's a little hard for us to, uh, to assemble burgers right now, but we're, we're getting there. Uh, burritos are another great one. So, uh, uh, everything except the rolling, of course, that's still a, still a challenge, but, uh, but we're really good at making burritos actually, uh, at poker, I'm not sure, uh, uh, you know, that's a pretty big trend these days for, uh, which is like a bowl with a sushi grade fish, uh, in rice, uh, you know, things like pasta or, um, uh, other soups, uh, things like that are also great.
Speaker 1 00:15:21 Uh, pizza again, there's lots of pizza robots out there and we're, uh, we're pretty good at putting toppings on pizza now. Um, uh, yeah, so really, uh, anything, uh, it's hard for us to find, uh, cuisines, which don't work very well, uh, for the robot things that are things that require a lot of really fine manipulation with your fingers are challenging, but everything else is really, uh, fair day.
Speaker 0 00:15:48 And so what are the main markets that you're pursuing? Like, I think of airports and malls and universities,
Speaker 1 00:15:54 Well, uh, everywhere that food is, is prepared. So we're really all about a robot in every restaurant and eventually a robot never kitchen, uh, but that's the, the, our long-term vision. Uh, but in the, in the near term where, uh, we're selling to, uh, corporate cafeterias, airports, hotels, uh, commissary kitchens, uh, ghost kitchens, you know, places where, uh, you, you make food for delivery only, um, quick service restaurants, uh, are really big ones. And then in, in sort of the lumped in there, right, that's, uh, in the cafeteria is right that schools, hospitals, uh, universities. Um, and so there's, there's a lot of, there's a huge number I would say of, of places that prepare, uh, prepare meals like this. And they're all great candidates for automation.
Speaker 0 00:16:49 And in the pre-call we talked about switching out your robots, right? Like you might have a, do something different in the morning at something different lunch and desserts in the afternoon or the evening.
Speaker 1 00:17:00 Exactly. So the, one of the great things about, uh, cobots, uh, is their small size and the flexibility of, uh, really of this kind of automation. And it's not. Um, and I think we're, we're seeing this in a lot of industries, right? Once you have a robot, you really want to find as many use cases for it as possible. And, and we are offering to our customers the, the opportunity to use the robot for a prep in the morning, then, uh, meal service, uh, during, during lunch. And then, uh, uh, like you said, other either other preparation or are, um, you know, preparing, uh, to go meals, uh, overnight for, uh, to sell the next morning, uh, right. There's a huge number of use cases, uh, when you have a, a robot to shaft to, uh, basically repeat the same task over and over again,
Speaker 0 00:17:56 And your vision, do you see robots in homes?
Speaker 1 00:17:59 Uh, absolutely. We're, you know, after all they say, build, build what you want. And, uh, I've, I'm dreaming of my, uh, my home, uh, robotic chef, uh, and, uh, that is definitely somewhere that we're, we're interested in going, uh, the cobots, the cobalt manufacturers are doing amazing things in, uh, uh, reducing costs and increasing capability every year. And we, uh, we're absolutely excited to follow on that trend and, um, uh, putting robots in the home when the price is appropriate.
Speaker 0 00:18:36 I think there's also a fun factor here too, where if I go to a Baskin Robbins that puts in these and in their ice cream parlors, I went just go there for that.
Speaker 1 00:18:47 Absolutely. Whenever we've taken this out, uh, you know, and show people a robot, we get lines, uh, you know, a block long. Uh, so it's, uh, absolutely people love watching robots and they're just, it is like a fascinating thing. And so that's, that's good for the first, uh, first couple of years, but, uh, really we're, we're thinking of this, like the, uh, the next microwave, right? So Alfred is, is really a tool and it's just that the, the first of its kind in the next generation of appliances,
Speaker 0 00:19:21 And for the average restaurant, we haven't talk about the ROI. How do you have a kind of a good idea for that?
Speaker 1 00:19:27 Uh, absolutely. So for the, um, uh, because it's robots as a service, the ROI to our customers is, is truly immediate. Uh, that's one of the really big benefits of that business model is that they can see a savings there, uh, on their bottom line, uh, in the first day.
Speaker 0 00:19:48 Nice. So winding up, what else do you do when you're not programming robots to cook?
Speaker 1 00:19:54 Uh, I'm also cooking myself, so I share, I share your hobby. There is a, is a bit of a foodie, uh, but, uh, uh, besides that, you know, I do enjoy, uh, playing outside with my family and lots of, uh, uh, rock climbing, what I can, uh, find the time, uh, you know, my romance or romance or real thought, uh, you know, it's a, it's a crazy time that we live in, right? You can, uh, uh, it's absolutely amazing what you could make the assistants do now.
Speaker 0 00:20:23 Yes, David's very exciting part. And I want to thank you. And how can people get ahold of you?
Speaker 1 00:20:27 Yeah, you should, uh, go to our website, uh, dexa.com, D E X a i.com. And there's a form there for hiring Alfred, get in touch
Speaker 0 00:20:39 And hire your robot. Our sponsor for this episode is Earhart automation systems. There are builds and commissions turnkey solutions for their worldwide clients. With over 80 years of precision manufacturing, they understand the complex world of robotics, automated manufacturing, and project management, delivering world-class custom automation on-time and on-budget contact one of their sales engineers to see what Earhart can build for you. I'd also like to thank and acknowledge our partner. the association for advancing automation. They're the leading automation trade association for robotics, vision and imaging motion control and motors and artificial intelligence technologies visit automate.org to more. And I'd like to thank our partner painted robot painted robot builds and integrates digital solutions. They're a web development firm that offers SEO, digital and social marketing, and can set up and connect CRM and other ERP tools to unify marketing sales and operations. And there are painted robot.com. If you'd like to get in touch with us at the robot industry podcast, you can find me Jim Beretta on LinkedIn. We'll see you next time. Thanks for listening. Be safe out there. Today's podcast was produced by customer attraction, industrial marketing, and I'd like to thank my nephew, Chris gray from music, Chris Coleman for audio support, my partner, Jen, and our partners, a three painted robot and our sponsor Earhart automation systems.
