Natan Linder on Tulip, Composability, and Why Lean Needs "Augmentation"
Manufacturing Happy Hour
On Manufacturing Happy Hour, Natan Linder, co-founder and CEO of Tulip Interfaces and chairman of Formlabs, explained what a "frontline operations platform" is and why he thinks manufacturing software has to change. His argument is that the people who design and run factories should have the same kind of modern, cloud-based, data-driven tools that knowledge workers already take for granted. Legacy "monolithic" systems, in his view, cannot provide them. The conversation covered his path from mobile software to 3D printing to Tulip, the idea of composability, his book Augmented Lean, his take on Industry 4.0, and what he expects factories of the future to look like.
"Fusing design and engineering" and the pull of the physical world
The host opened with a line from Linder's bio: that his work aims to fuse design and engineering to create novel human experiences. Linder said the phrase probably dates back to his MIT Media Lab application and has survived many revisions of his bio. He unpacked it by describing himself as an engineer who has spent his career building fairly complex hardware-and-software products, including mobile phones, robots, 3D printers, and other devices. Over the past decade or more, he noted, nothing built in "hardware land" is disconnected from the internet.
Put simply, he said, he loves it when people change how they work because of a technology and its interface. Hardware isn't always involved, but the work he likes usually touches the physical world. He pointed to the phone as an example. People's interaction with the internet once ran almost entirely through the browser, then it moved to the phone and "we never looked back." Phones used to be fairly dumb, though, and part of his career was spending a decade trying to make them smart and useful as interfaces. He described the MIT Media Lab as, among other things, "one of the largest toy stores on the East Coast," a place where you could build a lot of interesting interactions.
A decade in mobile, and no-code before it had a name
Asked how his time at companies like Samsung and Sun Microsystems shaped his view of frontline operations, Linder went back more than 20 years. Coming from an embedded background, he fell in love with mobile. Phones could run Java, build attractive UIs, and connect to the internet at the dawn of mobile connectivity. He was convinced this would change consumer life, but what fascinated him more was the prospect of it changing business. In his twenties he believed it was as big as PCs and Windows.
His early-2000s work focused on bringing software infrastructure to phones so that many developers could build for a very constrained platform. That meant low-power CPUs, little memory, and awkward interfaces with soft buttons that changed function. The core question was how to make lots of developers productive, and he connects that directly to Tulip's no-code approach: "I've been doing no code for 25 years," he said, and it started there. He also reasoned that if work was going to happen everywhere, on the go and at home, those devices would need to connect to business systems. His decade in mobile was largely interface work aimed at letting people build software quickly and repeatedly, spanning the pre- and post-iPhone and App Store eras.
When the host asked whether he personally foresaw this shift or the whole industry did, Linder said he was one of many practitioners of his generation who saw it. He compared it to how people talk about generative AI today, which he jokingly flagged as the conversation's first "buzzword bingo" hit. It changes everything and also doesn't. Word processing was transformed, since it's now cloud-based and collaborative, but it's still word processing. What he did not doubt back then was that the pace was accelerating, people would always be connected, and a huge number of business applications would need to be built.
Formlabs and Tulip as parallel projects
Linder explained that Formlabs grew out of a collaboration with students in his class at MIT as he was finishing his master's, while Tulip grew out of his PhD work. The two ran more or less in parallel. You can design a company as carefully as you like, he said, but you can't decide in advance which one will succeed.
Formlabs grew dramatically in its early years. It ran a very successful Kickstarter campaign and, in his words, pioneered the desktop stereolithography category. He sees the link between the two companies as "democratizing some piece of professional work." For Formlabs, that meant making this kind of technology routinely available to people who think and create in CAD, in terms of both capability and price. Doing that well required building the equipment, a supply chain for useful materials, and good software and user experience. Designers and engineers use the printers at the start of product development or on the shop floor to print jigs just in time when they reconfigure production. Once they had that access, he argued, they weren't going back. That, he said, is how Formlabs came to ship more than 100,000 printers. The company is about 12 years old and will turn 13 next September, by his count.
Tulip had been "a big idea" in his head for a long time during his research. He described facing "no-regret moments" that call for tough decisions. At the time, Formlabs was fairly stable with a leadership team suited to its stage, and he felt it was the right moment for Tulip to become a company rather than a project inside MIT. It wasn't an easy decision, he said, and "a lot gets figured out later." He remains Formlabs' chairman, which takes some time but doesn't keep him from building Tulip full-time.
What a frontline operations platform is
Asked to explain the category the way he would to a manufacturing executive over a drink, Linder described a frontline operations platform as a new way to rethink the industrial operations software that runs production. It delivers a platform as a service to the people who design, build, and operate production lines and other operational environments, such as assembly floors, machine shops, logistics centers, and quality labs, and to the people who work in them.
The tools he listed are cloud-native, no-code and low-code workflow building, ways to describe and model the operation and then run it in production, and built-in analytics on the data that flows out of the operation so it can be continuously improved. He contrasted this with the traditional stack, which rigidly separates a data acquisition layer, ERP integration, and a set of legacy tools in between, the MES, CMMS, and other acronyms. He called these static software that struggle to support what modern operations need, since the world runs on data and the ability to make decisions from it.
He framed it around the people doing the work. Both he and the host have been knowledge workers who take data-driven decisions with good tools for granted. "Why is that not the case in operations for everybody?" People demand it, he said, but the tools to achieve it are missing, and that gap is what the platform is meant to fill.
Why the category had to exist
The host noted that "platform" in manufacturing usually brings to mind IoT platforms or MES. Linder joked that he tries not to think about IoT platforms. He explained the need simply. Walk into almost any operational environment, whether a large multinational or a mid-market business, and "the internet as you and I know it is not there." People work with spreadsheets, paper, and "sticks and stones." He kept hearing the same story: there's no solution yet, we're on year three of implementing the ERP, we'll get to it later. Existing systems didn't quite fit what was needed, the value wasn't there, or the price didn't make sense.
This was before people knew how to pronounce "Industry 4.0," he said. He started crystallizing the thesis with Kubat, someone the host has also spent time with. When they talked to VCs and others, they were told some large company or global systems integrator already covered all of it and there was no reason for their proposed company to exist. Linder took that as a sign they were onto something. That was more than a decade ago, and he said they are "so far from done" even now.
Composability versus monoliths
The host asked what composability means in practice, given Tulip's contrast with "monolithic" legacy platforms. Linder said he didn't coin the term "monolith." It's simply a good description of the architecture. Traditionally, smart people in a room had to fix a finite scope to ship anything, especially before cloud software. Cloud delivery changed that. Good B2B enterprise software, he argued, should be built for continuous delivery. You log into Google Docs and find a new feature that works without anyone asking you. That isn't possible with boxed or on-premises enterprise software updated maybe once a year. Salesforce, Workday, and other large platforms serve huge numbers of professionals with continuous delivery, and he called it "just madness" to think the same won't happen in operations.
He described composability on two levels. At the IT architecture level, companies increasingly assemble best-of-breed components, one for connectivity, another for the data lake, and so on. Within a platform, he used Salesforce as an example. If the host were given access to Tulip's Salesforce instance, some things would be familiar because they're standard Salesforce, but other parts would be baffling without a guide, because Tulip has composed Salesforce to fit its business. Organizations have similar needs that still differ in practice. A composable production system, composable MES, or composable frontline operations platform, whatever you call it, puts that power in the hands of the people who design and build production environments. He said that is happening now, not someday.
Why physical operations especially need it
Linder argued the need is especially strong in operations because they are "complex and physical and dynamic." He pointed to a transient workforce with people constantly coming and going and needing training. He also pointed to how physical products move from new product introduction through engineering change management and product lifecycle, how products scale up, and supply chain management. As a result, two lines building the same product in the same company at two different sites may be slightly different, and the tools have to handle that.
He also brought in the IT side. He asked why OT should live apart from IT architecture, and why there should be long debates over specs and data models instead of letting data models be defined by the people doing the work. He said this doesn't dismiss the need for governance, which Tulip isn't neglecting, but companies and frontline workers need the ability to compose.
The software lens: lean, tools, and open source
The host summarized the theme as taking what we see in mobile and everyday technology and using it to make manufacturing more flexible. Linder agreed and offered software engineering as one clear lens. Early at Sun Microsystems, he recalled, a team lead dropped a giant binder describing a waterfall-style development model on his desk and told him not to read it but that they had to give it to him. Today software is built with agile, scrum, and sprints, which he said software engineers borrowed from lean.
The problem, as he described it, is that there aren't enough software engineers in the world, and especially not in operations, to write all the software flexible manufacturing requires. Engineers, being one himself, aren't lazy but like to be efficient and will automate whatever they can while keeping quality and throughput. That requires tools. The host asked whether more tools is how operations gets more software engineers, and Linder said yes. He added another angle: internet adoption and open-source adoption are correlated because software engineers share, and today's technology landscape is unimaginable without open source and open standards. Ecosystems create velocity. He asked where the open source and ecosystem for operations is, and said Tulip is busy building one with partners, with many others also having a role.
Augmented Lean: nothing missing, but something to add
In a quicker lightning round, the host asked what was missing from lean manufacturing that led Linder to write Augmented Lean. "Nothing is missing," he answered, which is why the book uses the word "augmentation," a term he has thought about for a long time given his PhD in augmented reality. Lean, from the Toyota Production System onward, laid great groundwork focused on organizations, and the book argues those principles are not wrong or broken. He mentioned that in a conversation with someone he refers to as Moak, that person told him they thought lean had so far failed, which he found very interesting.
The augmentation, he explained, comes from the fact that 30 years ago there was no internet, cloud, big data, or mobile phones, and three decades of rapid technology adoption haven't made it into the fundamentals of how lean works. Augmentation means giving lean practitioners the tools knowledge workers have. He described it as "almost a more radical version of lean" that gives them independence, because they control the creation of data and can make lean-driven decisions with it. He compared it to how, when an important deal changes in Salesforce, fifteen alerts and ten Slack messages fire in real time and everyone takes it for granted. Lean isn't broken, he said, but it needs augmentation to fit today's reality.
Industry 4.0, pilot purgatory, and COVID
The host said they had heard on Linder's Augmented Ops podcast that North America may have missed the chance to capitalize on Industry 4.0 while Europe did better, and asked how to avoid missing such opportunities in the future. Linder offered some history. Industry 4.0, "the European spelling," came out of a lot of EU research and project work, while the U.S. was "a bit more cowboy," implementing more and researching less. He considers the various labels, Industry 4.0, industrial IoT, industrial internet, individually insufficient and, taken together, potentially "wrong and dangerous," and not the most important thing. Technology to connect and collect data has existed under whatever name and has become cheaper and more available even in industrial settings.
He noted that while he criticized monoliths, there is good edge data-collection technology, and it will remain as long as there are PLCs. Some drivers will move to cloud-based connectivity. If the network talks to the cloud with good failover and the operation isn't high-frequency, that may be fine, though he wouldn't suggest running a petrochemical plant that way. He called these examples "different tastes" of how to think about the issue.
He then described COVID's effect. Before it, there was conservative reluctance to spend, "pilot purgatory," and fear of change. Then, more or less overnight, there were emergency product approvals, especially in pharmaceuticals and medical devices, and pressure to keep goods moving. "No one's questioning the cloud anymore," he said, because people at home needed to log into the factory. In a post-COVID world, you can't unlearn that.
His conclusion was that half the battle isn't technology but culture and adoption. He doesn't think the opportunity is lost; he thinks maturity has arrived, and the less we need the buzzwords, the more it shows readiness to accept that this is really happening. For companies, it's about staying competitive, an imperative rather than an option. He compared resisting it to launching a business with a digital strategy built on billboards, or companies that had fax machines and wanted to give email "a couple years."
Innovating in the factory first
The host asked when factories would start innovating first rather than importing technology from daily life. Linder said he already sees it among Tulip's partners in large and mid-market-plus companies. Tulip is less focused on 20-person mom-and-pop shops, which can use Tulip Essentials, and more on complex environments that make hard things. Those customers are already doing it, he said, because people on their shop floors are like anyone else. No one would start value stream mapping with a clipboard and paper. They'd open a visual tool, tie it to a spreadsheet, try to pull data from the PLC, and because they're engineers, build something. What they need help with is putting that into a platform that gives them what HR, marketing, and other functions have had for decades.
The factory of the future
Asked what the industries of the future will look like, Linder described highly connected, very composable architectures that blur the lines between edge and cloud and broaden the use cases they support, so the "front line" may sometimes extend into back-office functions such as HR. Being highly connected would allow "virtual value streams" in which one factory builds one thing and another builds another while being almost intrinsically linked, something he said they are already starting to see.
On economic value, he predicted companies adopting an agile, composable approach will see large returns in operational excellence, such as lower cost of poor quality, longer product lifecycles, and lower training costs. He said many of these outcomes are highly measurable and that they see them all the time.
Why a podcast and a book
Finally, the host asked how the Augmented Ops podcast and his book have helped Tulip. Linder said this kind of public role "is not natural to me at all," despite the host's remark that he seems natural. It was clear from the start, he said, that great companies help create real ecosystems, and Tulip wanted ways to bring a community together and tell its stories.
He was candid about the book's limits. It says many things but "doesn't give you the answer," because "we don't know the full answer just yet." It isn't yet time to declare what comes after Industry 4.0 or how to fix lean. Instead, he sees practitioners, including safety, lean, and quality engineers, factory heads, and executives redesigning supply chains, all grappling with the same issues from different angles. He described millions of people who identify with lean, quality, and operational excellence. They build factories, which he called essential to meeting humanity's needs from the most basic to leisure, and he expects factories to be there forever.
He said he believes a tectonic shift between eras is underway. Lean has had about 30 years, digital transformation about 30 years, and now "we're a minute before gen[erative AI] takes over." He said a community is needed to deal with that, and building it through Tulip and its ecosystem is what he is personally dedicated to.
Natan, you are someone I've wanted to have on Manufacturing Happy Hour for a while. I'm looking forward to this interview.
Thank you for having me. I'm glad to be here.
My first question for you is, I've read that your work aims to fuse design and engineering to create novel human experiences. That's a big statement. What do you mean by that?
I think this is from my bio and potentially my MIT Media Lab application, and it survived the many iterations of that bio. I haven't heard that one for a while, to be honest.
Okay.
But I guess if you go back, I'm an engineer. I've been building fairly complex hardware and software type products my entire career. Think mobile phones, robots, 3D printers, all sorts of devices. And usually when you build hardware, software, and more and more recently, but by recently I mean the past decade plus, nothing we make in hardware land is disconnected from the internet.
So when I'm talking about experiences, at the end of the day, to put it very simply, it's like I love when people change how they work as a result of technology and an interface. And it's not always the case that there is hardware involved, I'm saying it in general, but the stuff I like typically has something to do with the physical world.
Even the mobile phone, when you think about it. It used to be that all our interactions with information and perceiving the internet, up until some point, was using this weird new thing called the browser. What is that? But then there was the moment they moved to the phone, and obviously we never looked back, and now mobile is the internet, internet is mobile. But mobile phones used to be pretty dumb not too long ago, right? So also part of my career was spending a decade trying to make mobile phones smart and useful as an interface. So that's where that is really coming from.
And the MIT Media Lab is kind of like, many ways to describe it, it was definitely one of the largest toy stores on the East Coast.
Okay, never heard that described that way before.
So you can build a lot of really cool interactions, and so that's really where that characterization of the work I like doing, am excited doing, comes from.
Well, I'm excited to dive into your interest in the physical world, your background, but first I got to ask you a Manufacturing Happy Hour style question. So Tulip is a frontline operations platform. I haven't really heard that before until I heard of Tulip. So how do you describe what a frontline operations platform is as if you're having a beverage with someone?
Who is that someone?
Oh, I don't know.
Just so we calibrate, because it would be different, I guess.
Let's say an executive at a manufacturing company. I think that would be the persona.
Okay, so you want me to sell. There we go.
Basic, simple way to think about it: a frontline operations platform is a completely new way to rethink your industrial operations software that runs your production. At the broadest sense, it brings in a platform as a service to the people who are actually designing, building, operating production lines and operational environments. Those can be your assembly floors, machine shops, logistics centers, the lab where you do the quality, and to the people who live and work within those operations.
And it gives them tools, modern tools of the times we live in. So cloud native, workflow building with no code and low code, means to describe and characterize the operation and then put it into production or running it, giving you all the data that flows out of this operation into an internal analytics package such that you can continuously improve the operation.
And unlike the traditional stack, where you would have a very rigid definition of what is your data acquisition layer and what is, I don't know, the integration to your ERP, and maybe in the middle you have a bunch of tools, it could be legacy XYZs, where XYZ could be MES, CMMS, all sorts of tools with a bunch of acronyms that are supposed to give you that. Those are, I would call them, typical static software that are very hard to make support what modern operations need.
Which is, you know, I'm not going to belabor the point, but I think we can agree that the world runs on data today, and the ability to make decisions on the data. And that's what it does for the people who actually work in frontline operations, and we should keep this in mind. We talked about changing how people work, and for us, and I think you would agree, both of us in our careers were knowledge workers. So we take making decisions based on data with great tools as a given. Why is that not the case in operations for everybody? People demand that, but okay, where are the tools to actually achieve that? So that's what a frontline operations platform does.
And I'm optimistic we're getting closer to that. I know your company is helping people get closer to that, make that more the standard in the industry. Before we get further into talking about frontline operations, I want to hear a bit more about your background, because before you did Formlabs, before you did Tulip, you had experiences at companies like Samsung and Sun Microsystems. Tell us how those experiences influenced your perspectives around frontline operations.
Yeah, so I kind of touched on it a little bit. Back in the day, and this is like more than 20 years ago, I was in love with this technology called mobile. And I come from an embedded background, and I kind of saw what mobile phones are capable of, that they can run Java, and that was software that could create beautiful UIs, and they connected to the internet. And this is the dawn of the internet itself and the dawn of internet connectivity on mobile platforms.
And I was like, this is amazing, it's going to change everything we do in consumer. But I was fascinated with the fact that this is going to change business. So I was sure this is as big as PCs and Windows or something like that. And this is me in my 20s, and I was pretty ambitious, and I thought there's just a huge opportunity to do all sorts of software.
And a lot of the work there back in the day, I'm talking early 2000s, is bringing software infrastructure to those phones to help a lot of developers create software for that kind of platform. So constrained device, low power CPU, not a lot of memory, weird interfaces with soft buttons that used to change their functionality.
I remember that.
And all that. So that was like, wow, there's a lot of fundamental questions like how do you make a lot of developers be productive. So that kind of gets you back into, well, what is this no-code thing? I mean, I've been doing no code for 25 years. It started there.
And then it was a lot about, well, if that's the device, then it's going to change how organizations work, because work is going to start happening everywhere, on the go, at home. How do you connect it to business systems and things like that? So my decade in mobile was a lot of interface work. So what would make people quickly build software for this thing and actually repeatedly do it? And this is for the App Store and post the iPhone, so I kind of lived in an interesting era.
Yeah. What I'm curious about then: was it you that realized this was going to fundamentally change the way we were going to work, or was it the industry as a whole that realized it? I could see it going both ways.
I think I was just one of many people of my generation that were practitioners. And I think it's much like, people nowadays don't question that, you know, I'm sorry, this is the first time on the GenAI bingo, we're going to hit that virtual bell right now, maybe you can add it in post effects.
Sure.
It's like people growing up with GenAI right now, and they're like, oh my God, this changes everything. And it kind of does and it kind of doesn't, looking back. Because one of the examples I like: word processing completely changed, but it's still word processing. The main use case is still there, but now we do it on the cloud and collaboratively and all that.
So the point is, I don't know if I knew all the things back then, or my generation knew. Certainly I'm just one of many. But you couldn't ignore it back in the day, that the pace was really picking up and people are going to be always connected and there's a ton of business applications to be built.
One thing I'm curious about in your journey: you're at Tulip Interfaces now, you're still the chairman for Formlabs. That's the word I was looking for. There's an office right in Milwaukee where I live, I should have had it come off the tongue a lot quicker. But how did you go from mobile to 3D printing to what you're doing now at Tulip? I'm interested to hear how Formlabs fits into that journey.
So Formlabs was born out of a collaboration with the kids in my class at MIT when I was finishing my master's, and Tulip was also born out of my PhD work at MIT. And so they were like parallel tracks, and I was doing kind of two projects more or less at the same time. Only this thing is that you can design as much as you can for companies, but you can't really decide, oh, this one I'm going to design such that it's going to be successful over that period of time. That's impossible.
And Formlabs went through a very dramatic growth in its early years. It went through a very successful Kickstarter campaign. It pioneered the desktop stereolithography category. And the connection is about, sometimes we say at Tulip and other things, democratizing some piece of professional work. In Formlabs' case, it's making this type of technology routinely available to people who think and create in CAD. That's democratizing that access in terms of capabilities, price, and that's what defines access.
And to do that well, you need to build a piece of equipment, you need to create the right supply chain around it so you have a lot of useful materials, you need to have great software and user experience to enable that. And so designers and engineers, whether they're in the beginning of the product development funnel or they're down on the shop floor printing just in time because they need jigs to reconfigure their production, you're again back to the same thing: you're changing how they work.
And so would they go back to a world, now that they have this access to this technology, where they're not going to use 3D printing? That doesn't make sense to a lot of people. And that's how Formlabs grew to ship over 100,000 units of this printer, and then additional printers we developed. This company is almost, it's 12 years and going to be 13 next September, I believe. So that was a journey.
But at the same time, Tulip was a big idea I had in my head for a long time when I was working at MIT, and part of my research. And at some point there are these no-regret moments where you have to make tough decisions, and it was definitely not an easy one back in the day. It was a moment where I thought Formlabs is pretty stable, had a great leadership team for that moment in time, stage appropriate to the company, and the Tulip moment to be born as a company, outside of a project in the making within MIT, I kind of felt it.
And so it wasn't an easy decision, but sometimes you just do that, and a lot gets figured out later. That's sort of the gist of it. But I kind of never left. I'm not operating there, but being a chairman is a bunch of time, but it's not limiting me from full-time building Tulip.
Well, I think this is a nice segue to the next question, because when I think of platforms in our industry, I think of IoT platforms, I think of MES systems, manufacturing execution...
Try not to think about IoT platforms.
I can imagine why. So that's what I want to ask you next, because you're talking about how you had this passion for Tulip, it was always in the back of your mind. So why was there a need for a frontline operations category then?
It's really simple. You would walk into any one of those operational environments I described earlier, this could be in a huge multinational or a much smaller mid-market business, and the internet as you and I know it is not there, and people are working with spreadsheets, papers, sticks and stones. And you keep hearing the same thing: we don't have a solution, we're on year three of implementing the ERP, we'll get there later. So they're telling you that story on lack of adoption of technology, and it always comes to, we're here, but it doesn't exactly do what we need, we'll come back to that later.
And the other thing they're saying is the value is not there or the price doesn't make sense. And by the way, again in our buzzword bingo, this is even before people knew how to pronounce Industry 4.0. The rise and decline of the use of Industry 4.0 came after I kind of started crystallizing the thesis with Rony Kubat, who I know you spent time with now.
We were thinking this is nuts, and then talking to all the usual suspects, VCs and others, and everybody told us, oh, so-and-so large-scale company has it all, or this global SI, they build it all, there's no reason for this company you're talking about to exist. This is before we even founded the company. And I'm like, okay, great, that means we're onto something. And that was the case more than a decade ago, and we're so far from done even a decade in.
I've got a question about that a little later on. I want to ask you, talk to me about composability a little bit. I saw that out on the floor, and the way I look at it is you've moved from having what you call monoliths, these platforms that you couldn't do much with, it's like, this is the way it's set up. So what does composability mean in practice?
Yeah, by the way, we didn't coin this. I'm not the only person on the planet calling legacy software monoliths. It's just a good way to describe their architecture. A bunch of people, likely very smart, motivated, sat in a room, and in software development you have to be finite, otherwise you will never get to deliver code to anybody. Certainly that was the case before cloud software emerged, but now we're solidly in that space, so it changed how we deliver software.
So software in general, B2B enterprise, the good ones in my view, should be designed with continuous delivery. So for example, when you delightfully log into your Google Docs or whatever other vendor you're using, you're like, oh wow, there is this new feature, and it works for you, and nobody asks you if you want this, and you're like, okay, that's great. That's continuous delivery, and we take that for granted as well. And you can't do that in this old architecture where you're bound to software in the box, or enterprise software that's being updated maybe once a year, maybe sometimes much more.
It's a good comparison.
And that doesn't make any sense, because so many B2B platform tools, think Salesforce, think Workday, and all these kinds of large-scale platforms that support hundreds of thousands if not millions and millions of professionals in their organizations, work with continuous delivery, and they were born on the cloud. And to think that that's not
…going to happen operation is just madness from my perspective. So where composability comes into place: first of all, we have to zoom out and understand composability is also not unique just for operations. We are a to of bringing it to operations, but in IT you're seeing best-of-breed type of IT architectures composing all sorts of different — we're going to use this for our connectivity, and we're going to use that for our data lake, and we're going to use this and that. That is like one level to think about composability. But when you dive into the platforms themselves — like, have you ever used Salesforce before? I have. So if I give you access to our Salesforce, there's some stuff you will be able to do because it's Salesforce, and you will go into the standard interfaces and you'll find this report, that report or whatever. But some stuff you'll have no idea about unless someone takes you through why, because we've composed Salesforce to fit the needs of our business.
And this is back to the point I was saying before, that it's heterogeneous in nature. It's like we have the same needs, but they're kind of different between different organizations, even though they're very, very similar. And composable production systems, or composable MES, or composable frontline operations — you name it the way you see fit — gives that power to the people who are actually designing and building production environments. And it's not a maybe one day, perhaps; it's now.
And the reason this is happening, or the reason I think the need for composability is so pronounced in these environments, is because they're complex and physical and dynamic. That makes it very hard. So what do I mean by that? Think about the transitionary workforce: how many people come in and out and need to train. Think about how real physical products are being born and engineered, from NPI — so new product introduction — to engineering change management, to product life cycle, so how long does a product scale up, to supply chain management. All that is physical stuff. That means that if both of us have the same production line that even builds the same product in the same company across two different sites, they might be slightly different. And to do that, you need tools that are able to deal with this kind of intensive set of differences.
So that's just on the face value of it. And then a lot of people like to talk about the IT side of things, and that's also where it comes into play, because why does your OT need to live sort of independent of your IT architecture? It shouldn't. And why do we need to have long debates on what are the right specs, or the data model — this conundrum, what's the data model — and why can't it be defined by the people actually doing the work? Again, not to belittle the need for governance and things like that; that's not something we're neglecting in our conversations. But this ability to compose — companies need that, and people who do the work need that. So that's where I think it's at at the moment.
Well, I mean, if I'm looking for a theme in this conversation so far, I feel like you're taking the things you've seen in the mobile world, that we see in our everyday life, and you're making manufacturing more flexible, to oversimplify it a bit.
I don't think you're over— I think you're on the right track. One very clear vector you can think this through is software, the lens of software, and the art and craft of software engineering. When I started — and we talked about Sun Microsystems — I remember one of the team leads, and I'm this super young engineer, and they bring me this giant binder with like itle e model of developing software, sticking it on my table: "Don't read this, but I've got to give you this." And this is like the waterfall deluxe version. And of course, how do we develop software today? It's agile, scrums, sprints, all these. And where is all this coming from? From lean. Yeah. Software engineers and software architects stole all that, or grabbed it over from lean, because software is very complex.
And we know that software is sort of the glue and the conduit in which we get the data and that kind of stuff. You know what's the problem? There's not enough software engineers in the world, period. Specifically, there's not enough of them in operations to create all the software that you need to have this flexible manufacturing that you're talking about. And by the way, if you're doing software, software engineers are kind of — I don't want to, I am one, so I'm being a little bit careful — it's not that we're lazy, it's just we like to be efficient. I get you. A good engineer is a lazy engineer, I've heard that. We automation, and we're going to use whatever we can to get by and to still get the quality and throughput and all that kind of stuff, and that means tools. Where are those tools?
And by the way, is that how we get more software engineers in operations, is more tools?
Yes. And here's another angle to think about this. If you plot the adoption of the internet, as pronounced by browsers, active sessions, this and that, and open-source adoption, they're correlated. Why are they correlated? Because software engineers share, and open source is a thing. You cannot imagine the technological landscape today without open source and open standards. So ecosystem matters, and that creates velocity. Where is open source and ecosystem for operations? You're creating it, is what I'm looking at. We're busy creating it with a bunch of great companies and partners, and yeah, we have a role to play and many other people have a role to play. So all of that is what makes manufacturing flexible down the line.
You brought up lean manufacturing, and I've heard you kind of do lightning rounds in your interviews around buzzword bingo. Since we've got a little less than 10 minutes left in our interview, I do want to get your quick answers on a few categories, starting with lean manufacturing. So you have a book called Augmented Lean. So what was missing from lean manufacturing — like, we know about eliminating waste, increasing productivity — that the world needed Augmented Lean to exist?
So nothing is missing. Okay. And that's why we use the word augmentation. Augmentation has a lot of — like, I spent my PhD around augmented reality, and this word augmentation, I've been thinking about it for a long time from a technological perspective. So it's about the term, or the field that it defines, where lean operations or lean manufacturing, lean companies — this is from the Toyota Production System onwards, as it's been captured and then proliferated — was fundamentally focused on the organizations. It was laying out great groundwork, and I don't think those principles are — and that's what we're saying in the book — they're not wrong, and they're not broken, even though in my conversation with moak that started at some point, he told me, "I think lean has failed so far," which was really, really interesting, how he said that.
But the augmentation comes from the simple fact that 30 years ago there's no internet, there's no cloud, there's no big data, there's no mobile phones. And you can't ignore the fact that these 30 years of accelerated technology adoption did not proliferate into the fundamentals of how lean works, of how you empower and you create sort of superhumans. Because now when you think about the augmentation, we talked about the knowledge worker, so how do you give those tools to lean practitioners? It's almost a more radical version of lean that gives them independence, because they're in command of the data, the creation of the data, such as their ability to make decisions that are lean-driven. Or just the way we're living today, where something happens in some important deal you're tracking with Salesforce — what happens? Like 15 alerts fire almost instantly, 10 Slacks and all that kind of stuff, and it's all real time, and you're like, of course. So lean is not broken, in my view, but it needs augmentation to the reality that we live in today.
So I've got three more questions for you: two related to industry, one something completely different. Okay. You brought up Industry 4.0 earlier, and I'm curious, as far as capitalizing on digital transformation, I feel like this Industry 4.0 era has been going on for what, maybe a decade now or so? Decade plus, yeah. And I was listening to one of your podcasts on Augmented Ops when I was running around Boston Common this morning, and it sounds like we might have missed the opportunity to capitalize in there, at least in North America, where maybe Europe did a bit better job of capitalizing on it. I'm curious to hear your perspectives on it, because judging by your facial expressions I might not be getting this right. But the main question I have is, what do we do to not miss opportunities like this in the future?
So just a little bit of historical context. The project of Industry 4.0 — the European spelling, there's a lot of "I" at the end — so there's a lot of European Union work and project research that went into that, and the US were just a bit more cowboys, who kind of implement stuff, less research, I'd say. And I think there were a lot of different names to describe different aspects of similar phenomena, and they're all kind of insufficient and, put together, potentially wrong and dangerous. So you call this Industry 4.0, and this is Industrial IoT, or industrial internet, and all these things, quite honestly, they're not the most important things, because technology to connect and collect data — doesn't matter how you call it — existed and has become more available and more commoditized, even for industrial settings.
So we're seeing this phenomenon specifically where we talked about monoliths, but on the edge there's some great edge data collection technology, and it will be as long as there are PLCs and the like. And guess what, some of the drivers will move to cloud-based connectivity, and if your network talks to the cloud and you have good failovers and you're not in high-frequency land, you might be okay with that. If that's your operation — I wouldn't suggest you run a petrochemical plant on something like that — then maybe that's okay.
And then COVID happens. I'm just giving you a little bit of different tastes of thinking about this. And then, from a very traditional, almost conservative "let's not spend money on all these kinds of things," and pilot purgatory — I'm sure you heard that name, we hear it a lot — and people are afraid of change, overnight, more or less, lots of emergency product approvals, certainly in pharmaceuticals and med device, but also in other places, because you had to get trucks with whatever to the — the economy was kind of hit with COVID. And no one's questioning the cloud anymore. Why? Because you're at home and you need to log into the factory and you need to see what's going on. And so now we're in a post-COVID world. How can we unlearn what we've learned? We can't.
So long story short, 50% of the battle is not technology, it's the culture and the adoption of these different sets of technology. So I don't think the opportunity is lost. I actually think maturity is here. The less we need those buzzwords, the more it tells you how ready we are to accept that, okay, this is actually happening. And you know why this is important for companies, and potentially the listeners, our customers, our partners? It's about staying competitive. So it's an imperative. It's kind of like I'm telling you, look, I'm going to start this great business, and my whole digital strategy — we're going to work with billboards, I don't need this digital advertising. The world has changed, and the earlier companies accept that and put themselves to it, the faster they will get back on the competitive track. And I think it's similar to the way some companies were like, you know, we have fax machines, so this email thing, let's give it a couple years.
Sounds like what we need to do is find a way — because a lot of the things you've been saying are, we're learning, we're seeing cloud in our daily lives, we're seeing technology in our daily lives, and we're bringing it to the factory. I'm wondering when we're going to start innovating in the factory first a little bit.
I mean, we're seeing it all the time, and this is what I'm trying to say. For us, it's exhibited in the behavior of our partners in large companies and mid-market-plus companies, where we're less supporting the 20-person mom-and-pop shop — I mean, they can also use Tulip Essentials as much as they want, but that's not our core customer. Core customers are complex environments that make hard stuff. They're already doing it. And why? Because the humans on their shop floors are just like you and me. We would not start our work with a clipboard with a piece of paper on it and say that's how we should do value stream mapping. You would likely fire up Miro or whatever visual tool, and then we'll tie it to a spreadsheet, and then we'll try, okay, we need to get some data from the PLC, and maybe there's a — because we're engineers, so we'll build something like that. But that is exactly where the customers need help: to put it in a platform that gives them what their fellows and any other constituency — HR, marketing, whatever — already have had for decades.
So I know we're at the end of our conversation here, so you've got to give me one-minute answers to each of these last questions, and one of them's a big one. What will the future, or the industries of the future, look like based on our conversation today?
Highly connected, very composable architectures that blur the lines between the edge and cloud and IT, as well as the set of use cases that they're supporting. So not just "this is the system that supports the front line" — and the front line will exist also sometimes in the back end, maybe in the HR department. Okay, and I like that. And highly connected means that you can have virtual value streams, where one factory is building one thing and another factory is building another thing, and they're almost intrinsically connected. We're already starting to see that. So that is a bright composable future we're headed towards. From the economic value of that, it's going to be tremendously impactful. Companies that adopt this sort of agile, composable approach will see huge dividends in operational excellence that will translate into lower P, lower cost of poor quality, higher product life cycles, lower cost of training people. There's a lot of highly measurable things that we're seeing all the time. So that's the future I'm seeing.
Well, I think that's a nice summary around our conversation. The final question is, you've got quite the personality in the manufacturing and automation world. You do your own podcast, you've written a book — I feel like more people need to be doing that. So how has having Augmented Ops, your book, how has that helped you in Tulip Interfaces?
So first of all, this is not natural to me at all. Okay, you seem like a natural. I don't feel like a natural. Fair. Okay, but it was very clear from the beginning — and I think it's true for many great companies — that they help to create real ecosystems. And I…
I think we're on it, and we wanted to have different ways to bring the community together and tell the stories, and that's why we were pushing on it. Because if you look at the book, the book is saying many things, but it doesn't give you the answer. And the reason is because we don't know the full answer just yet. It's not the time to say, okay, what happens after Industry 4.0, and here is how you fix lean ABC.
We're more or less in a point in time where we're seeing the practitioners and the people they serve. So think about the engineers, it could be safety, lean, quality; they can be the head of factory or the executive working on the new supply chain design. We're all grappling with those issues from different aspects, so we wanted to get the word out there. We wanted a place where we can discuss and help facilitate this community.
There are millions of people who self-identify as lean, quality and operational excellence people. They built the factories. You know why factories are important? Because they're actually helping the human race survive. Yeah, lifeblood of our economy. Helping our economy, it's our most basic needs and all the way to our leisure and everything in between, and they will be there forever.
And I really think there is a sort of a tectonic shift in the eras right now. If you think lean has been 30 years and digital transformation has been 30 years and all that kind of stuff, and now we're a minute before gen takes over, and we're, you know what, so we have to have a community to deal with that. And that's what I'm personally dedicated to do, and through Tulip and our ecosystem want to usher.
Well, and you're part of it too. I'm trying. Sharing stories, getting out there, having conversations as if you're having a drink with someone. So I appreciate all the work you're doing. Like I said, you do come off as a natural, so clearly all the work you're doing sharing the story is paying off. So Natan, I just want to thank you so much for jumping on Manufacturing Happy Hour.
Thank you for having me, and thank you for visiting Tulip. Exciting to have you. It was a great tour, great conversation. We'll see you soon. All right, we'll see you soon. Cheers. Cheers.
Article published