This week on the podcast (augmentedpod), futurist Trond Undheim interviews Erik Mirandette, Head of Product and Ecosystem at Tulip. This is episode 75 of Season 2: "Designing a Worker Friendly Industrial System."
In this conversation, we talk about what designing and redesigning a worker-friendly industrial system might entail, how to build capabilities and not point solutions that simply fix existing use cases to empower operators and workers along the way.
My takeaway: It is unusual to hear the case for manufacturing efficiency, coupled with worker empowerment, and then expressed so clearly as a systems dynamics problem that needs to have an overall fix instead of just attacking.
Thanks for listening. If you liked the show, subscribe at Augmented podcast.co or in your preferred podcast player and rate us with five stars. If you liked this episode, you might also like episode 73 "The Challenge of Frontline Operations."
Augmented reveals the stories behind the new era of industrial operations, where technology will restore the agility of frontline workers. Technology is changing rapidly. What’s next in the digital factory? Who is leading the change? What are the key skills to learn and how to stay up to date on manufacturing and industry 4.0? Augmented is a podcast for industrial leaders, process engineers, and shop floor operators, hosted by futurist Trond Arne Undheim (@trondau), and presented by Tulip, the frontline operations platform.
The Augmented podcast is created in association with Tulip, the frontline operations platform that connects the people, machines, devices, and systems used in a production or logistics process in a physical location. Tulip is democratizing technology and empowering those closest to operations to solve problems. Tulip is also hiring. You can find Tulip at Tulip.co
Please share this show with colleagues who care about where industrial tech is heading.
To find us on social media is easy, we are Augmented Pod on LinkedIn and Twitter, and Augmented Podcast on Facebook and YouTube:
See you next time. Augmented--industrial conversations that matter.
Trond Arne Undheim: Welcome to another episode of the Augmented podcast. Augmented reveals the stories behind the new era of industrial operations where technology will restore the agility of frontline workers. Augmented brings in industrial conversations that matter.
In episode 75 of the Augmented Podcast, the topic is: "Designing a Worker Friendly Industrial System." Our guest is Erik Mirandette, Head of Product and Ecosystem at Tulip.
In this conversation, we talk about what designing and redesigning a worker-friendly industrial system might entail, how to build capabilities and not point solutions that simply fix existing use cases, instead empowering operators and workers along the way.
What matters is, do I have Google maps or ways or whatever? And the point is it's not the skill of the individual as much as it's like the accessibility of the state of that system at any given point in time. Now traffic is kind of an easiest problem to solve because at the end of the day, all agents are trying to go from point a to point B in the most expedient manner possible, right?
[00:00:21] Like everybody's really solving for the same problem. And so a common UI and a common set of capabilities in a single. How to build capabilities and not point solutions that simply fix existing use cases and instead empower operators and workers along the way. Augmented covers the future of manufacturing.
[00:00:38] Today, we interview thought leaders across the manufacturing industry on how they manage operational challenges. Our vision is a world where technology restores. Oh, frontline workers. Augmented is a podcast for industry leaders hosted by futurists throne. Undheim presented by Tulip.co the frontline [00:01:00] operations platform.
[00:01:00] Eric, I'm thrilled to have you here. Yeah, no, I appreciate you
[00:01:03] Erik Mirandette: having me. I'm looking forward to this.
[00:01:05] Trond Arne Undheim: Eric you and I know each other a little bit, and I thought I knew you in the sense that people always think they know another person because there's familiarity. And you've seen them over a certain period of time.
[00:01:15] But I wanted to introduce you by saying Eric was a high school athletes. He went to the us air force academy and you have a business degree. You had a stint in venture capital, and now you work at to an industrial software startup. That's a pretty reasonable explanation of your life or is it
[00:01:35] Erik Mirandette: well, there's a bit more to it than that.
[00:01:36] It was a somewhat nonlinear path into. VC for that matter, which was a relatively brief stint about a year or so, and then eventually into tulip. But I went, as you said to the air force academy out of high school, this was about 2001 when I was getting started there. And then, a country was headed toward war.
[00:01:55] I was a sophomore at the air force academy at the time. You know, when I realized that I I had [00:02:00] very little knowledge of who was out there, how the world worked, who these people were that, we were going to war against at the time. And I wasn't in principle of posts to it necessarily, but it was more like, I think an awareness of my ignorance at the time I ended up anyway, this sort of led to me leaving the air force academy for a leave of absence.
[00:02:20] And moving to north Africa to work with a nonprofit organization where I was mostly in Morocco and for about a year and a half, I lived with an Arab family, and I worked a couple of different things. Medical assistance to west African refugees was the first thing that I focused on spent the better part of a year, getting food and medication.
[00:02:41] To a group of migrants that were across the Sahara and trying to get into Europe. Then there was a big earthquake in north Africa at the time. This was back in 2004, and I moved to the effected area to the, the epicenter of this earthquake and spent the remainder of my time there doing earthquake reconstruction efforts had a team of about 15 local nationals that I [00:03:00] was responsible for, and then following my time with the non-prime.
[00:03:04] Before I was headed back to the air force academy. I decided to take a trip. So I had been working with these west African refugees for the better part of a year and a half. And their stories were just so compelling and so different than anything I had ever been exposed to before. So I decided to take a trip and I wrote a dirt bike from Cape town, South Africa, up to Cairo Egypt.
[00:03:24] Trond Arne Undheim: You wrote a book about it, which I'll just read the only road north. It's a very compelling and tragic story. Why did you feel like you needed to write about this? Well, I guess
[00:03:37] Erik Mirandette: to fast forward a bit, when I got to Egypt, I got caught up in a very nasty suicide bombing that ended up taking the life of my brother.
[00:03:46] And, I never intended to write a book per se. It was more of, for me, just a means of catharsis and trying to make sense of. Some things that are not easy to make sense of. So I took to the pen as I often do, and trying [00:04:00] to make sense of these things in, at the end of the day, you know, that was a story that I wanted to share.
[00:04:05] And so I ended up contracting with the public. The habit published. So that's the book, the only road north.
[00:04:11] Trond Arne Undheim: Yeah. I mean, you backpack through Africa, you had many interesting experiences to meeting so many people along the way. And the first part of the story is sort of the dream for any young person, I guess, going out on their own and really discovering themselves in the face of the adversity of there not really being a road, you said the only way north, but you are traversing places and.
[00:04:37] Border crossings and meeting a lot of interesting people along the way. I'm just curious, just a little bit about that story. When you think back about that time, is it possible for you to separate the end of that story from the big. In other words, what you set out thinking that you would learn I'm imagining was something about other people are leading a very interesting and [00:05:00] different life.
[00:05:00] And I now know all about it. And then suddenly at the end of it, it turned out, perhaps you're learning a lot more about yourself. That's at least how I read your story.
[00:05:10] Erik Mirandette: Yeah, sure. Yeah. Going into it, young idealistic, like most young men, there's nothing in the world that can have. You can do all things, obviously life-changing a set of experiences for many reasons, frankly, I suppose maybe I should fast forward a little bit to answer the question.
[00:05:26] Obviously dramatically impacted my worldview. I did go back to the air force academy. I ended up taking a third year. To recover physically. That's when it was during that year, I ended up moving to Hawaii and I bought a one way ticket with $500 to my name. And I ended up getting a job tending bar on the north shore of Kwai, a place where I think frankly, lots of people go to kind of get lost.
[00:05:47] And it was during that time that I wrote the only road north. And then I did go back to the air force academy as was my intention all along though my focus in world. Shifted dramatically, as you can imagine, I went to the air force academy, thinking I would fly planes and, wanted [00:06:00] to be an astronaut.
[00:06:00] I majored in physics and astronautical engineering. I went back to the air force academy and shifted my studies to political science and philosophy. And then instead of going the pilot route, I went into intelligence operations because another thing at the time, this was 2005, 2006. So this was at the peak of the first surgeon Iraq.
[00:06:20] And the other thing that occurred to me is. Yeah, I spoke Arabic. I had a lot of friends who spoke Arabic and these are not bad people. My wife is Lebanese and Muslim. Right. And I think the thing that was important for me at that point in my life was know I wanted to be not just on the execution side of things where you're just fulfilling your orders, but I actually want it to be.
[00:06:42] Helping inform the apparatus that made these types of decisions. So I went into intelligence operations. I focused in human intelligence operations specifically, and that's what I did for the next, seven years. And that took me all over the world, ironically, to no place that actually spoke Arabic, which is perhaps a [00:07:00] representative irony frankly, of our military machine.
[00:07:03] But I'm very proud of the. I do think I, I had an opportunity to spend a lot of time on the ground, getting to know these folks, and that's what I did for a chapter of my life. And then came a point in time where that was done and it was time to move on and do something else. And that's where I suppose the the story picks back up, me transitioning out of the military to MIT for grad school and in the business school.
[00:07:24] Trond Arne Undheim: So then you entered the business space as a different person and in a somewhat different way than, I guess you had imagined. Tell me a little bit about what made you join Tulip.co and what is your understanding of what Tulip.co is building? You know, we talked about designing a worker friendly industrial system in the prep for this call.
[00:07:44] Tell me a little bit about what you mean by that. And was that already something you fully understood it to be when you accepted the. Yeah.
[00:07:53] Erik Mirandette: Well, I mean, when I joined too, if I don't even know that there was a job to accept, it was a pretty early team and we didn't have any of the things we have today in terms of [00:08:00] product or funding or market traction or customers for that matter.
[00:08:03] I mean, it was pretty early stage, but the journey to Tulip.co is perhaps an interesting one I got out of the military really was no earthly idea. I had gotten relatively good at subverting foreign governments. No idea. How in the world that it can be applied to a civilian context. No idea what the heck I was going to be doing.
[00:08:19] But I did a brief sentence BC, as I said, and I fell in love with the early stage space. These entrepreneurs who were taking on hard problems and operating with a small team against the odds and mid an incredible degree of ambiguity. That felt very familiar to me and very energizing to me.
[00:08:34] I much prefer to be where the action is rather than on the sidelines. And so I decided to come to MIT for grad school with the intention of finding a place in that space. That's where I first semester of grad school is when I met Aton and Ronnie Tulip.co had just recently come out of lab. And that's what led to the initial connection there.
[00:08:52] Now, why Tulip.co. There's many factors that influence it. But I think like one of the things that I've found to be most compelling, if we talk about the problem that Tulip.co is [00:09:00] solving, at the end of the day, our goal, our vision, our mission is to provide the people who run the world's frontline operations.
[00:09:07] They fundamentally better set of tools that allow them to do their job more effectively, full stop. Now, practically speaking, what does that mean? You know, we're talking about environments in which things are being made most typically. So the phone in your pocket, the shirt on your back, the computer, through which we're talking now and the device through which whoever listens to this will be listening.
[00:09:28] All of this stuff is like made physically by people somewhere. Right? And if you step foot into where these things are made, It is striking how underserved by technology these people are. And now there've been leaps and bounds with robotics and automation, lots of advances there. But if you look at like the technology that actually serves the people that exist within that run these operations.
[00:09:50] It's been largely untouched by technology in the last 20 to 30 years. It's as if the internet has not yet happened in these environments. And that disparity was just striking [00:10:00] to me. First of all. Now, if we think about the context in which these problems are solved, really what we're talking about now is manufacturing or more broadly frontline operations to include like lab environments and warehousing environments and things like that.
[00:10:13] Things that aren't strictly defined as manufacturers, but a similar space. Yeah, manufacturing it. Stacks are not something that's intrinsically interesting to. Or at least they weren't, I've actually gotten quite a bit more into it, but what was compelling to me is the sheer scope and scale of the problem.
[00:10:28] Right? So if you think about the enormity of this, I get literally everything that's that, that we rely on in our everyday life was built somewhere by someone collectively, these folks represent. About 25% of the global labor force. There's about 650 million workers or so that are employed by these various operations accounts for north of 20% of global GDP.
[00:10:48] It's massive on a scale that almost boggles the mind. There was an element of intrinsic interest associated with the sheer scope and scale of the problem that Tulip.co endeavored to solve. But on a more profound level for [00:11:00] me is also the conditions in which these folks work. I grew up in grand rapids, Michigan, my family I've got I come from a big Catholic family.
[00:11:07] I've got a lot of cousins. I couldn't tell you the number, but I would, I think it's fair to say 15 to 20% of my cousins are employed in manufacturing, operations, right. System integrators, or they work on the floor. And these are not folks who run these factories. They're the folks who work in these factories.
[00:11:21] You know, they're in the union. And when I came into sort of the ivory tower of academia, and I heard leaders of these industrial operations, talking about how do you increase efficiency and solve these problems? There's often talk of minimize the role of these folks. Like the 650 million workers that I mentioned it was, we'd automate them if we could, but we can't.
[00:11:42] So now we just have to like, remove their ability to think and make decisions. You hear talk like. It's just fundamentally wrong, and one of the things that was attractive to me about Tulip.co is from the very beginning, there was a respect for, and a celebration of an elevation. The people who are in the center of these [00:12:00] operations as an imperative to improving these operations.
[00:12:03] I found that to be a worthy cause on a personal level. And we can talk a little bit about why the tulip had that position. It's not, the social end wasn't necessarily the driving impaired. In the early days, it was more it's like while it is true, but it's also the fact that automation just fundamentally can't solve the problems that people can solve.
[00:12:21] Computers are very good at some things and people are very good at very different things. And if you look where the lab that tulip came out of was the fluid interfaces lab. And so this is all about removing the friction between the human and the computer and leveraging each for their respective strengths is the driving force behind it, and I think back to at the time, it was very much not invoked.
[00:12:40] Take this position in the early days of Tulip.co. This is about the same time. If you'll recall, when Elon Musk was getting ready to ramp model three production, and there was a comment he made that I grabbed headlines, which was he said, I want to walk up to the side of the factory and I want a big dial.
[00:12:56] And when I turn that dial, I want model threes to come out faster. The [00:13:00] implication being, this is a lights out facility. Fully automated. If you recall how that went, fast forward a year into production, hell is what the headlines described. It, Tesla was not doing well. Model three production was not going well.
[00:13:13] And the wash out of that was we over automated. We automated things that don't make sense, things that can, that machines are not like things like upholstery operations. And also the reality that like any of these operations, these manufacturing operations, They're complex systems which is to say the, in specifically they're complex adaptive systems complex is to say that no one actually knows what's happening across the entirety of the system at any given point in time.
[00:13:38] It's not something that can be understood. And there's all sorts of like nonlinear outcomes and, little things happen that lead to unpredictable outcomes, right? It's a complex adaptive system in the sense that these systems are populated with age. That are constantly learning and changing their behavior in response to changing conditions around them.
[00:13:57] And when I say agents, I'm specifically referring to [00:14:00] these frontline workers, the people in these operations, and that's a very interesting type of system. If you think about trying to remove all decision-making from that system, it leads to certain problems. In many conditions, there are edge cases in which it can be done, but it doesn't describe the state of most of these systems.
[00:14:17] The way that you optimize these systems is by. Creating feedback loops and making the sharing of information available in easy. So allowing people to solve problems very quickly to look does this through self-serve tools that we make available in our product, but letting people solve their problem while also giving them access to the state of the system around.
[00:14:40] Another example of a complex adaptive system that I think is appropriate is a, traffic is a pretty good one. There are many others, biological systems and such, but traffic is an interesting one. If you think about traffic, if I want to go from Boston to New York, it almost doesn't matter how good of a driver.
[00:14:56] What matters is, do I have Google maps or ways or whatever? [00:15:00] And the point is it's not the skill of the individual as much as it's like the accessibility of the state of that system at any given point in time. Now traffic is an easiest problem to solve because at the end of the day, all agents are trying to go from point a to point B in the most expedient manner possible, right?
[00:15:16] Like everybody's really solving for the same problem. And so a common UI and a common set of capabilities in a single. Can sufficiently solve that problem for all agents in that system. If you take that and you apply that to a manufacturing operation, the things that people are trying to do are much more highly varied, which means that a single application doesn't solve all of these problems.
[00:15:39] What you really need is the ability to specify the type of application to solve the specific problem that you're addressing. But everyone needs to be able to do this, like the concept of making this ability available to all the agents in the system. And thereby creating a shared awareness or transparency of the state of the system across all of these agents is how you would take this problem on
[00:15:59] Trond Arne Undheim: a [00:16:00] lot of what you speak about is building a new system.
[00:16:04] I don't know what you would say, but if you look at the history of industry and automation, it was a very. And arguably actually up until this day has been a very centralized system. You shared with me and analogy from like the beginning of time, there were physical reasons why these systems couldn't be decentralized.
[00:16:21] Right? So there, there are certain structures. What, powering a factory was not simple, for example, right before electricity. So there was a physical dictate that I guess that's how the concept of a station originated. Because you have to stay in your station, not because it's nice to have for everyone to have a station.
[00:16:39] It's because you need to drive power to that station. You need to drive power to whatever activity you're supposed to carry out. Now that logic seems to have outlasted. The initial problem. What you're talking about is actually a mind shift, right? Because not only is factory work becoming somewhat more physically decentralized.
[00:16:59] You know, [00:17:00] globally, but also individually and actually down to the specific thing that someone is doing at any given moment, because that's, this is what you're talking about. These apps, they can change. Any one person can use many apps or it can change what they're doing, but still coordinate with the overall system.
[00:17:17] Tell us a little bit about how complicated that is to do, because to an uninitiated, you might just think, oh, all of this is software. It's like super easy it's it happens at the speed of light and there are no. And we have come so far with algorithms and this should be very easy to do. Something tells me that it hasn't been that.
[00:17:35] Erik Mirandette: Yeah. We just moved into the new office space here and a part of Boston called assembly row. So named, because this is where they used to assemble for automobiles back in the fifties, they've adopted the theme across this whole industrial park and retail space and such in I was walking into one of the buildings one day and there was this massive 20 foot tall.
[00:17:54] 10 foot wide cylinder, and I was like, wondering what in the world is this thing, right? What, where did this come from? [00:18:00] And it turns out that comes from an old manufacturing facility. This is back, as you said, when there was centralized power generation. So these were. Steam powered or sometimes water powered, when you would take a central source of power, convert that to mechanical energy and then distribute that mechanical energy to the various workstations across the whole of the facility using typically like a belt in shaft system.
[00:18:23] And this was, this particular piece of wood was a gear belt. It was shifted gears of the system. What that means is that if you think about how you build one of those facilities, You need to know what you're building, how to build it. Like what parts are required about how many that you will be building and where every single physical station is going to be in this and this facility really, before you lay the first brick.
[00:18:45] Right? So these are like multiple years, tens of years. When you conceive the system before you get your first product, and then they're very inflexible in an unchanging world, no big deal. That's fine. Like we have a thing, it produces the product all good. As you said, the thing that changed here [00:19:00] fundamentally is.
[00:19:01] Decentralized power generation. And now every CNC and every lady and every thing that you use in a modern manufacturing facility has its own onboard power source. You can just plug it into the wall and you can do whatever you need to do. And you can configure these environments in a much more flexible and reconfigurable way.
[00:19:17] And there's no going back. You couldn't imagine. Any manufacturer today saying, you know what? That central power generation idea was a good one. We should go back and we should do that again. Like it's, there's no going back. You never would. You never could. And it strikes me that if you look at the way, the same companies today build their it stack, they build an instrument, the systems that capture information and feed information to the people responsible for these operations.
[00:19:43] It's highly analogous. You build the system. To solve for a very specific problem. This problem, you start by collecting the requirements and then saying, okay, let's spec the system. Okay. Let's go ahead and agree on a vendor. Okay. Let's go ahead and customize this monolithic software system with a rigid backend data [00:20:00] model to meet the needs of this specific process that's been defined at this point 12 months earlier, or so.
[00:20:05] Then we're going to actually implement it. We're going to roll it out. And once it's there, guess what? You're never changing it. It's too inflexible. It's too expensive. You're never going to change it, which is, just not how software works anymore. And furthermore, it doesn't need to work that way anymore.
[00:20:19] Now, like not only is it out odds with the reality, like product a sells product B doesn't who saw that coming, or you run into, and now we're talking about these nonlinear systems, right? You have a supply chain shortage, let's say there's a global pandemic or a ship gets stuck in the Suez canal or something like that.
[00:20:33] These are, Purely hypothetical, obviously, but like these things happen and you're left without the ability to respond to them. And that is largely due to the it decisions that are made years in advance and there's no need for it anymore. And increasingly we're seeing our customer base say, we're never going back and there's an alternative.
[00:20:52] And obviously tulip is built with this as a sort of a foremost consideration.
[00:20:56] Trond Arne Undheim: Yeah, and it is fascinating. So on this podcast we call it [00:21:00] Augmented and arguably argumentation, you're building out something on top of a system, but the argument here is you're building it with the human as the main system.
[00:21:10] And then you were supporting that system by other types of technologies is actually a pretty drastic departure, from the way industry has been thinking, not just in terms of infrastructure, but also arguably in terms of managing. So you're out there explaining the way our product works and obviously arguing that it matters but more importantly, as customers are onboarding and starting to use it, I'm imagining that your role is a little bit also to Trudeau broaden their view and help them along to try to envision how they could use.
[00:21:42] These opportunities now, what are some of the other properties of kind of this new reality of not being physically tethered? So you talked about some design considerations that are, opening up the layers. So self-serve is a principle. No code is another one, right? So the reason it is self-service, that's a [00:22:00] good part of the code.
[00:22:01] Modularized and he's ready as like hot pluggable components overall. What are the features that matter in this, if we are going to endeavor to, to create this new world where the worker is more empowered and matters more on the operator, I guess who's organizing this type of work has a new set of tools.
[00:22:20] Tell us a little bit more about what these tools look like, and what's gone into building.
[00:22:25] Erik Mirandette: There's two fundamental sets of considerations. The first is architecture. And then the second is form. If we talk about architecture, this thing called virtualization happened, whereas previously, if you wanted to run software, you had to have physical hardware available to run that software and what AWS and Azure have done, they've virtualized these capabilities, which means that anybody that can connect to the internet has the full capabilities of.
[00:22:52] Best infrastructure systems in the world available to them without having to incur the overhead of maintaining these [00:23:00] facilities. Now how various companies decide and what goes in the cloud and what goes on prem varies, but overwhelmingly it's a fundamental differentiator from what existed 10 years or so prior, and to.
[00:23:13] You are seeing overwhelmingly a migration to web native cloud native software offerings. And the reason for this is that flexibility component, right? You can scale up or scale down as much as you need to without having to procure and maintain additional hardware to do it right. Dramatically lowering the cost and increasing the flexibility to the customer.
[00:23:32] So that's the architecture. Tulip.co in case I haven't stated it explicitly is cloud native. So we run these odd services and we do have an edge component, which is a hardware that we can talk about perhaps.
[00:23:45] Trond Arne Undheim: So before we go to the forum, I just had a question here because someone told me 95% really is the same tech everywhere.
[00:23:52] Right? So you said you, you put it in a form of cloud, but you could also say internet or web type technologies, and then you had cloud and then [00:24:00] now edge and the constraints and opportunities of edge on these devices and stuff, but then 5% or whatever percentage it is customized.
[00:24:08] It's only. Mind boggling to think about let's even though so many of these components are arguably the same all across industry and use cases, whatever percentage that's not the same has to be. So I guess customized in a very specific way. Why is that the case? And is that the case in a different way in manufacturing and more heavy industries than it is in any other end?
[00:24:32] Erik Mirandette: Yeah. So we've addressed the architecture. How do we make the capabilities available now we're shifting gears and talking about the capabilities themselves and how did we do and why did we build it in the way that we did? The reality is that every shop floor is different. It's like history.
[00:24:45] It's not, it doesn't repeat itself, but it sure rhymes, it's, they're all, it's all, they're all a little bit different, but fundamentally they're doing similar things. These differences, it turns out matter quite a bit. These differences are the reasons. If we think about the process of implementing a rigid system, if they were all the same, [00:25:00] it wouldn't matter.
[00:25:00] You build the perfect system once and then you roll it out. But because they're all a little bit different, there's that need for customization and then the resulting in flexibility that makes the existing solutions ill-suited solve this problem in a good way. It's the reason there's the need that there is today.
[00:25:14] So let's talk about the form first. Tulip.co we've built a no-code manufacturing application platform. The key point here is that we build at the platform level. We don't build at the application level. What that means is that we've basically made the ability to build purpose specific applications that solve very specific problems available to the people who are in these systems, the frontline users, the engineers, the supervisors, the directors.
[00:25:38] Now it's most frequently the case that someone has solved this problem before. Using a very similar method, think about all of your standard lean practices. And in such a case, you don't have to start from scratch. We've got a library full of best practices that are available on the internet.
[00:25:54] Talking to the architecture, you can go to our website, you can find the one that works for your purposes. You can click a button and you can [00:26:00] start with a solution that's anywhere from 50% to 90% done and you can get started there and that's going to solve a number of these problems, but there is enough variables.
[00:26:11] That there is the need for customization and configuration, even within these applications that are mostly good to go, which is why building up the platform layer and giving these frontline users the ability to configure and adopt and adjust, and then as their processes evolve and as the nature of the problems they're solving also.
[00:26:28] These solutions can evolve with them. You know, you can import these applications get started there. You can configure them, solve the problem today, but then as your operation changes, again, to it as a platform you can using this sort of Lego like approach, you can continue to add on functionality or adjust functionality as required based on your sort of operational considerations.
[00:26:47] Trond Arne Undheim: I wanted to ask you one question, because I get this a lot. When we start talking about augmentation, people immediately jump to augmented reality, which is a whole thing these days, [00:27:00] right? Because of discussions around metaverse and other things, and what it alludes to, but certainly for a while now, it has.
[00:27:06] Device focus. So it's also a form factor focused, but it's focused on whatever happens to be the latest device that then brings, a digital rendering on top of your reality to lip seems to have deemphasized. Whatever the end point device is, and I'm assuming Tulip.co supports many such devices and we'll support them to the extent that they become important, on the shop floor or wherever a client wants to have it, but it, why is it the.
[00:27:35] Tulip.co seems somewhat agnostic at the device level. You said Tulip.co focuses on the platform level and then brings apps so that operators can make their own apps and essentially make use of apps that others have made. But why isn't that last leg? The, so the Augmented devices. Why isn't that a major focus, many other companies focus a lot on these endpoint [00:28:00] devices that sort of arguably are the last mile for the worker to make those improvements.
[00:28:05] Erik Mirandette: Vince Lombardi said football is about blocking and tackling, right. There are a lot of capabilities out there and there's a lot of things you could do. Value is created when processes improved and inefficiencies are eliminated, right? There are cases in which an augmented reality may be a perfect platform to improve process, but it is not an end in of itself.
[00:28:27] Tulips focuses on helping our customers. Find inefficiency in their process and then eliminate that inefficiency. We do run native on any device that you can connect to the internet. Basically these could be smartphones, they could be tablets that could be touch screens, which constitute the majority of our deployed endpoints, but it could be a real warehouse headset or Vuzix, or, a wearable of some sort, and we run on those as well.
[00:28:51] It's a web native tech technology. You can put these applications and any device that can connect to the internet. The reason that we're device agnostic is because that's not the primary [00:29:00] consideration. The primary consideration is on making sure that we solve problems related to process. And if those are the appropriate platforms to do that, we're all about it.
[00:29:08] And you'll find them well supported. I can tell you that very few people are at that state today. There's a lot of hype about it. There's not a whole lot of them that are being run in production.
[00:29:19] Trond Arne Undheim: Exactly. At the end of the day, two leap is about enabling production. So the moment, tens of thousands of devices of whatever sort require that sort of support my guess is, would support it.
[00:29:29] But we
[00:29:30] Erik Mirandette: do support it to be clear. We'd like you, if you come into the office today, you can see a demo. We do support it today. It's on our library. If you go to our library page, you can download these applications and get started with them. We just aren't emphasizing.
[00:29:41] Trond Arne Undheim: Right. But there seems to be a point why that's not being emphasized and B being device agnostic is a point of Tulum is making.
[00:29:49] And I'm just trying to dig a little deeper into that because I guess it's quite easy to tell yourself to a device because it becomes very tangible. Why don't we speak a little bit about the kinds [00:30:00] of efficiencies that Tulip.co has enabled. I'm just curious to understand your take on what are the main things that this client.
[00:30:09] New type of system designed for an industry or frontline operational tasks. What are the kinds of things that can be improved by using such an architect?
[00:30:19] Erik Mirandette: Sure, no, everything we've been talking about today is very abstract. Let's make it very concrete for a moment in a very tangible environment. What does that look like and feel like, and I'll give you an example of one of our early customers, one of the early engagements that I was in remained deeply involved with.
[00:30:34] We had a quality engineer at an apparel manufacturer. They're one of our early customers. He was responsible for the first wave of the implementation of Tulip.co to up at the time there was, now there's relatively broad and people know what it is and people have they know why there's a need for it at the time.
[00:30:48] This wasn't the case. And so he, he asks questions like, what is this thing? And why do I need it? And how can I get advantage of, and I said Tell me what you do. Day-to-day and what are the problems that you're solving? Oh, and it was with the caveat of these are manufacturing people.
[00:30:59] There's [00:31:00] not a high tolerance for BS. Right. So he's oh, and by the way, I don't have time for this. I said what what do you do? Like, how can we work on this? And he said walk me through his process and said, every Friday, I spend the better part of a whole day entering data into Excel.
[00:31:12] I said where does what data are you it's as quality defect data that he's entering into Excel? I said where does the quality defect data come from? And we walk out onto the shop floor. He says, it comes from these forums as well. How many of these forums are there? He's there's 10 lines and there's five stations at each line.
[00:31:25] And every shift does one and then I catch them on Friday. And this is batch manufacturing by the way, this is apparel they're making. I was like, okay, so you spend everyday, you collect all these forms and then you spend your Fridays entering this into an Excel. And then you get at the end of that, you get a graph about what your defect rate was.
[00:31:40] And then you maybe get some information about what the root cause of this defect was, but it's like way too late in the game to do anything about it, because you just made 5,000 of that shoe and your defect rate was whatever, 7% the damage has been done. And I convinced, I said, give us one. Don't enter the data into an Excel, give us that Friday.
[00:31:57] And what we did is we built a simple application and we, the end [00:32:00] of that day, we replaced all of those quality defect forms with a simple application. That was the exact same as the form. Instead of writing it on a piece of paper, it said, enter entered into this form and now. The next Monday when people came back in and started working, suddenly these forms didn't need to be collected because all of this data was available real time.
[00:32:17] The operators like it, because it was easier for them, it was less, they had to worry about. And it's also, using the same technology that they have in their pockets. These are not dumb people. They know how to use computers. Maybe they're not formally trained engineers, but guess what?
[00:32:28] They all have smartphones in their pockets. So the operators. And then, we start getting this information real time, and then the question becomes, now you've got a Friday free, right? So what are we going to do with this Friday? What we did is we built a second application that I said, like of the root cause that we're seeing here being piped in through these applications.
[00:32:44] Like where does that happen in the value stream? He said this root cause happens here. And that accounts for about 60%. This root cause happens there, that accounts for another 20% of it so on and so forth. So what we did is we built a feedback loop into that line. So rather than [00:33:00] waiting until the batch of 5,000 chews or whatever was done, when the first defect made it to the quality inspection station, there was a real-time feedback loop that was introduced to the person that was actually responsible for that defect.
[00:33:12] And they were able to alter their behavior and prevent additional defects from being made. So that's one very concrete example. How this set of capabilities can be used in the context of one of these systems to identify and eliminate inefficiencies and defects that are related to how process is being conducted.
[00:33:32] Trond Arne Undheim: Eric, I only have a last question for you. You seem to be describing the emergence of. Almost self-correcting system with a caveat that we've been talking about, a human centric self-correcting system. This is not some sort of automated situation where you're perfecting a machine. You are perfecting a system where human workers are involved and operators are perfecting it.
[00:33:55] And then you're using machines as part of that process. Where can this and [00:34:00] where will it go? So there's, Tulip.co, there's a few other players here adopting these kinds of approaches as you're looking into. The immediate future of manufacturing, the immediate future of all of frontline operations and practices that have these kinds of corrective behaviors that presumably can be improved.
[00:34:19] Where is all this leading us? And in what
[00:34:22] Erik Mirandette: time. I think first and foremost, the message here is that people want to do a good job. And if you give them the tools, they'll do a good job, and that's like core to the philosophy of tulip. And so really what tulip is all about is making sure people have access to these tools.
[00:34:36] And again, we're respecting the individuals and celebrating the individuals who run these operations, fully empowering them to be there. They're most effective is the goal. Now, where is this? And in what timeframe, if you think back to the total addressable market for these capabilities, anyone who made anything ever basically is what we're talking about?
[00:34:53] We've seen a lot of success recently, we've grown quite a bit year over year for each of the last several years. If you look at the number of customers [00:35:00] we have today, they measure in the several hundreds with operations and something like 18 countries at last count.
[00:35:05] Thousands of users, this represents as a proportion of the total addressable market is infant testimony. So there are many orders of magnitude. More people that can benefit from these technologies. Then there are people who are currently benefiting from the technologies. And so much of what we're doing now is really oriented around, making this accessible to those folks.
[00:35:26] Now, some of that is, go to market commercial, but really what I'm referring to is building that scale into the product. So scalability of the product, when we first built this tool, you think about that engagement I had with the apparel manufacturer at the time. If you had a hundred people all using Tulip.co at the same time in that company, that was a really big customer for us at the time.
[00:35:49] This was several years ago. Now, today we have customers saying, Hey, we're going to be deploying a thousand plus stations over the next quarter. And we'll have hundreds of [00:36:00] customers of which is true. And so there's the exponential scale component to it. And making sure that the product from a UI perspective.
[00:36:07] From an organizational management perspective, which is to say, how do we just organize the sheer volume of content that this requires in the data that's being consumed and making that the analytics and the insights available to folks as the total amount of information that's being captured increases, how do we continue to make that available to folks, but then also from a governance and permissions perspective, how do we make sure that this can be consumed at that new level of scale?
[00:36:29] That's a really big product focus for us. Another product focus for us is, we. You've mentioned no-code a couple of times. So have I really, we don't think about tulip is necessarily having to be no code. Really what we think about tulip as one of our first product principles is we want to build tulip with low barriers, but high ceilings, which means anybody who wants to get in to Tulip.co.
[00:36:52] If you are comfortable with PowerPoint, then you've spent some time with Excel. You can probably get started and be pretty effective building and deploying [00:37:00] applications. Now that said. We also have software professionals that work within it, departments that are responsible for connecting Tulip.co to third-party systems are want the ability to customize certain components.
[00:37:11] There's a place for them as well. And there's a lot of focus on making sure that. Stay true to both of those principles, making it easy for the new user to get value out of, to it quickly, but then also not limiting those power users who can do so much more.
[00:37:25] Trond Arne Undheim: Well, I have to say the scope is astounding, but the degree of individual.
[00:37:31] Kind of empowerment that comes coupled with it. I guess it's that balance. That is so amazing with what you're talking about, because you're talking about building a system that caters to individual, initiative wherever they are in whatever station they work and whatever problem they are facing, but also one that can be governed and organized so that a large organization can trust it.
[00:37:54] And a. That would seem to me to be in a live environment with many moving physical [00:38:00] parts, some of which are dangerous because they're on the industry shop floor. It does seem to me now quite different from building any other software tool where you agree it is the kinds of system constraints that you've been elaborating.
[00:38:12] Make it a very different proposition from just building an app. Yeah.
[00:38:16] Erik Mirandette: Yeah. We sometimes joke, we're not building a dog walking app, if it's a much trickier problem set. Yeah. And you know what, like last point there is, to that end, we're empowering our users and I'll tell you what the emergent behaviors we see in the use cases that our customers take on with.
[00:38:32] Oftentimes far exceed things that we had ever imagined. I would say that there is a level of creativity and ingenuity that is exhibited, that we did not initially conceive when we would build a certain capability into the end of the platform to see what people do with these capabilities as they apply it to their reality far exceeds anything that we would be able to centrally plan or create for them.
[00:38:55] Right. My point is, If the burden was on us to create the end state [00:39:00] application, we would be missing out on so much creativity that we see across our customer base.
[00:39:06] Trond Arne Undheim: Well, Eric, thanks you so much for sharing your journey into Tulip.co and the journey that is awaiting people who are, I guess, brave enough to go on these journeys that are now about to happen across frontline operations, because it's a landscape that is rapidly gonna evolve.
[00:39:24] It seems thank you.
[00:39:24] Erik Mirandette: That's really my pleasure. Trond thanks for having me.
[00:39:28] Trond Arne Undheim: You have just listened to episode 75 of the Augmented podcast with hosts thrown on a Undheim. The topic was designing a worker friendly industrial system, and our guest was Eric Miranda, head of product and ecosystem at Tulip.co.
[00:39:43] In this conversation, we talked about redesigning industrial systems to empower operators and workers along the way. And my takeaway that is that it is unusual to hear the case for manufacturing efficiency, coupled with worker empowerment, and then expressed [00:40:00] so clearly as a systems dynamics problem that needs to have an overall fix instead of just attacking.
[00:40:07] Thanks for listening. If you liked the show, subscribe at Augmented podcast.co or in your preferred podcast player and rate us with five stars. If you liked this episode, you might also like episode 73 on the challenge of frontline operations, Augmented industrial conversations that matter.
Head of Product and Ecosystem
Erik Mirandette leads Product and Ecosystem at Tulip, the leader in frontline operations technology spun out of MIT. He has 15+ years of experience in building and leading times in operations and product development. Today, Tulip has raised over $150M dollars in venture capital, employs 200+ people, and supports 250+ enterprise customers with operations in over 20+ countries.
Prior to Tulip, Erik served as a military officer for over six years assigned to the Strategic Counterintelligence Branch with tours in Afghanistan, Southeast Asia, and Japan. He was awarded both the Army and the Air Force Commendation Medals for meritorious service in combat. Erik, a Tillman Scholar, also spent time in the non-profit space based in North Africa, and in venture capital prior to joining Tulip.
Erik holds a B.S. from the Air Force Academy, an M.A from Norwich University, an M.B.A from M.I.T. He is the author of “The Only Road North” a memoir of his experiences in Africa.