Marc: We have a product that has 2, 000 components. It took 7, 000 steps to produce one machine, and we were able to digitalize the entire process in under five months. You don't need a large bankroll or a team of 50 or 100 software engineers. You need a few passionate process engineers and manufacturing engineers that are excited about new technology. It doesn't take a lot to get going. Announcer: You're listening to Augmented Ops, where manufacturing meets innovation. We highlight the transformative ideas and technologies shaping the front lines of operations, helping you stay ahead of the curve in the rapidly evolving world of industrial tech. Here's your host, Natan Linder. CEO and co founder of Tulip, the frontline operations platform. Natan: Welcome back to Augmented Ops. Today, we're switching things up from our usual formula. Your host for this episode is Liz Reynolds. Liz and I recorded an episode back in July called Reindustrializing America, which I highly suggest you check out. Most recently, she served on the National Economic Council, a special assistant to the president for manufacturing and economic development. She worked on topics like supply chain resilience, national manufacturing strategy, and the broader industrial policy agenda. Now she works as an MIT professor, as well as a strategic advisor here at Tulip. With that, I will let Liz take it away. Liz: Welcome back to Augmented Ops. We've heard on this podcast and elsewhere that we are in a new industrial revolution. This industrial revolution or industrial transformation is driven by both new advanced manufacturing technologies and production systems. That's the digitalization of the production system with digital tools, whether it's robotics, 3D printing, AI, et cetera, but also by new priorities, which have been driven in large part by changes on the global landscape. So resilience and resilient supply chains, the pandemic laid bare real vulnerabilities that are affecting both companies and countries. And so the question of how do we build resilient supply chains and how and where companies should manufacture is now front and center. Another key topic is sustainability. The threats of climate change and catastrophic weather events are causing massive damage and destruction, along with, of course, an existential threat to the planet. So companies and countries, again, examining ways to reduce their carbon footprint and generate more efficient, renewable energy. And both of these very much are advanced by digitalization. So again, Introduction of all these new technologies that bring the data analytics into the process so that we can find lots of improvements and ways to improve quality, safety, productivity, et cetera, in our resilient supply chains, in our decarbonization efforts. And without that digitalization, we really can't achieve the previous two points. So in today's podcast, we're gonna delve into one case of how advanced manufacturing processes are helping a medical device startup. disrupt the dialysis industry. This is a great case of seeing how new ways of manufacturing are key to a company's ability to compete and disrupt an established industry. It's a great example as well of how the adoption of digital technologies creates competitive advantage for small and medium sized firms so they can not only compete globally but also be disruptive and leaders in established industries as well as new frontier industries. It's also a great example of how this kind of technology is human centric, where the worker's being augmented by the technology as at the center of this revolution, rather than trying to figure out ways to automate the worker out of the equation. I think also, for me, it's interesting. A great example of how we need to, quote, digitalize the middle. This is the small and medium sized firms that are the backbone of U. S. manufacturing. And it really needs to be a public priority for U. S. manufacturing, particularly given the investments and priorities around semiconductors, clean energy, the defense industrial base, a whole host of ways in which we need to bring this digital capability to our small and medium sized firms. And with that, let me welcome Marc Nash to Augmented Ops. Marc is the SVP of Operations and R& D at Outset Medical. Hi, Marc. Great to see you. Marc: Hey, Liz. Nice to be here. Thanks for having me on the show. Liz: Yes. Well, excited for this conversation. Let me first start with asking you whether you could provide a little bit about your background. Marc: Oh yeah, of course. Happy to. So I've been in the medical device and operations and R& D management for the past 18 years. Spent a lot of time both in the US and abroad. Working in both greenfield facilities, brownfield facilities, around four different countries and three continents, and my focus has really been around value generation, value creation, and looking into the areas of manufacturing about how we can capture that value quicker, in a more systematic approach. And as we've kind of recently pivoted over to the fourth industrial revolution, This is really exciting times as we talk about digitalization and what that can do for an enterprise or business. Liz: Well, you've got a great perspective on what's happened over time, particularly in the space of so called industrial 4. 0. Can you introduce Outset and give a little history of the company? Marc: Of course. Outset was founded in 2010. We're in the business of providing people with end stage renal disease. and acute renal dysfunction with a technology advanced solution that helps restore patient identity and dignity. So we're really focused on hemodialysis. It's a critical life sustaining therapy that many people around 880, 000 people in the US alone required to receive treatment at least three, if not four times a week. And we're in the space of developing new age technologies for both the piece of capital equipment, as well as the consumable. Like what was the piece of equipment to allow for people to, in essence, do hemodialysis in the comfort of their home? Liz: That feels like or sounds like a very revolutionary type of step in the dialysis industry. Marc: Yeah, it's a big one. It's um, you know, for far too long, there's been a duopoly, not only in the US, but kind of globally in the area of dialysis. For about 40 years in the U. S. alone, there have been two major players in this space. And with such few, you know, new emerging technologies and an entrance into the, into the landscape, it's kind of led to a complacency around technology. That the devices that are out there are good enough, that these patients don't deserve better. Imagine having to go to a clinic three or four days a week, and you're spending approximately four to five hours a time at the clinic. Plus the travel time you're talking about. 20 to 25 hours a week, almost a full time job, if you will, of just trying to maintain your health, maintain your dialysis schedule. That can be a really big burden for both the patient as well as the family and the caregiver. Imagine if you could reduce that time by almost 50%, imagine if you could in the comfort of your house with a caregiver, your significant other, your brother, your sister or church member, be able to perform your dialysis treatment at your schedule when you need it and have the physicians and nursing staff at the local hospital or clinic able to monitor your treatment remotely. That's a game changer, and that's what Outset has been doing, uh, since 2010. Liz: Those sounds like really significant innovations for the patient and for the industry. So you're really essentially trying to disrupt a well established industry with two dominant players. Sounds like a number of other verticals we might be able to point to. So what in your mind sets the Outset device apart from what's out there today, and, and how is your manufacturing process, in particular, part of what you see as your competitive advantage? Marc: So, since the beginning of Outset in 2010, we've always had the same mission. To put our patients first, to put them in the center of how we develop and ultimately produce our product. If you were to look at our machine, it kind of reminds you more of a machine you would have at home for a consumer device, rather than a piece of very necessary medical equipment. So from the design of how it looks, how it functions in fields and how the user interacts with the device is really part of our competitive advantage. A lot of the devices you see in the Marcet and we've been around for decades, Contains a lot of very complex tubing sets and procedures and protocols that a nurse has to follow, which makes it not feasible to be used in a home setting. Many of the devices on the Marcet don't contain a lot of sensor technologies that allow the provider or the patient to understand how the treatment is going. So when we started to design Tablo, that's the device that Outset produces, We started to think about the patient in the center of the room, if you will, of the product. The device contains over 70 different sensing technologies and devices. We can measure a plethora of pressures along the entire device. We can measure blood flow rates and a multitude of other parameters. which not only provides out, and we'll go into this later in the discussion, with very, very useful data about how to make the machine more improved and with better efficiencies, but it also provides the caretaking team with special data that allows them to ensure that the treatment was as successful as possible. Liz: So, When we think about the process that you've set up, you're going up against, it's a little bit of a David and Goliath situation, and you've got to set up something that's going to be better, stronger, faster than the existing model. What was your approach when you thought about the actual manufacturing process? Marc: So similar to our product, we needed to find a solution. That was creative and unique to our current industry and our competition. As you mentioned, we needed a manufacturing process that was faster, more robust, more agile, and that could scale with our business. So some of the things we were looking for as we started to think about that solution was, you know, if we know that we have to scale quickly, And we know that we have to be better than our competition. We also needed to be able to glean insights from a, from a quality perspective faster. That was one of the number one things we were concerned. We also wanted to make sure that as we knew that our product would be well received in the Marcet, we wanted to be able to ensure that we could flex the facility and the manufacturing of the different devices and products we produce in a way that aligned with the business needs. So if certain devices and accessories had a higher demand at certain periods of the year, We would wanna be able to flex, you know, and reduce as necessary. We also wanted to make sure that as we think about our device having two-way connectivity, we wanted to make sure as we collect all this data from the device, we can seamlessly integrate that data of the device with our manufacturing data. 'cause if ultimately we can create that digital thread, that digital fabric. We could really gleam a lot of insights into reliability, uh, supplier quality, manufacturing quality, which would be important as we try to drive additional cost savings and efficiency in outdoor use. Liz: So that sounds like the classic integration of IT and OT, which is, you know, the holy grail at this point. So what did this mean in terms, I know you, you know, basically started from the outset with a digital manufacturing platform that is basically driving data. Constantly with workers on the frontline and throughout the operations. So what were sort of the, the impact of that, or how did you see that play out over time? Marc: That's kind of the pinnacle of our story, right? It's, it's always great to talk about how we could have a vision or we could put something into play, but ultimately it's the execution of those horizon results that, that truly matter, right? We're really happy to say that we've seen significant improvements. Uh, we saw improvements in our ability to detect defects and finish goods and reduce those defects by 70%. As you are mentioning, we're continually getting data, so imagine if you have a world where as you collect data on how your finished testing is going, you can connect those results directly to the results that we're seeing upstream in your inline testing processes and your workmanship standards. And if we see that there's a discrepancy between how our inlay test process detected anomalies and our finished good test process, we can make improvements upstream. So we're driving improvements upstream, ultimately reducing costs and necessary reworking and finished goods. Using the digital platform, Tulip, we've really been able to streamline our training processes. All of our apps and all of the work instructions that we provide in Tulip, in essence, look very similar from one application and one work instruction to the next. This allows for the training to be reduced by over 50 percent because the collaborators, what we call our operators, they have a easy to understand user interface that guides them through the process step by step. And ultimately, you know, as we think about the world of lean manufacturing and always trying to reduce waste and non value added work, we've reduced our transactional work on our manufacturing floor by 90%. Well, typically in manufacturing organizations, you see a mini army of individuals Kind of doing work order transactions in the background. They're moving materials from the warehouse to the manufacturing floor, from one location on the floor to the next. At outset, one device has 2, 000 components. Imagine if you're producing hundreds of devices a month, you can have tens of thousands of work orders per day or per week that need to be transacted. It doesn't really benefit the end user. So how did we streamline that? Well, we made API enterprise level connectivity between our manufacturing execution system and Tulip. And our ERP, or our quality management system, or a plethora of other enterprise level systems that we've connected to date. 2, Liz: 000 components is daunting. And it feels like at the core of what you're talking about is lean approach and an understanding in the importance of lean. But what you've said to me is that actually, you'd be much more effective at lean When you have the digital platform, because you're actually helping drive the data, which drives your attention to the anomalies, which then drives you toward continuous improvement, etc. Can you talk about that interface a little bit between Lean and the digital? Marc: Sure. So Lean started back in the, you know, 50s and 60s, and it was a great way to discover the seven wastes and start to provide root cause analysis on how you could clear up those wastes. But one thing that Lean always required was data. And a lot of time, it would take weeks or months to collect data, so when people talk about continuous improvement, in the traditional sense, it's actually really step improvement, right? They collect data for two or three months, they analyze that data, let's say, within a Kaizen or, you know, a burst of activities to go create a solution over a week or two. And then they go implement. So that flywheel, if you will, is a three to six month endeavor and you're continually stepping. But what if that data was available in real time every single day? And that data also had alerts that told you when your process is deviating from its normal trend, both in a good way and a bad way. What if you're becoming more efficient? That's also important to understand how that's happening. So with having that data at our fingertips on a day to day basis, we're able to incrementally make day to day improvements that ultimately has drive significant benefit, as I kind of had mentioned before with some statistics. Liz: Yeah, it's just an incredible sort of, you know, 21st century story. of what we've had since Toyota introduced in many ways, Lean and our continuous improvement models and, you know, all that we've learned over the decades. And now what does that look like in a augmented lean context? How about the two way device connectivity that you've talked about and how you are harnessing sort of the IIoT architecture. Can you talk a little bit about that bi directional Connectivity, what that actually looks like. I think in part, you know, there's a democratization of this data in some way because of this connectivity and bidirectional process. Marc: Agreed. So if we think about Tablo, one of the unique features that makes it really a game changer in the Marcet is bi directional connectivity. So what is bi directional or two way connectivity? It means that we can send data to the device, right? That data could be a software release, a feature set, a patch for improvement. But we also collect a plethora of data from the machine, both physiological data on blood pressure rates, heartbeat, etc. As well as a lot of data around device performance. Microsoft The sensors that I was mentioning earlier, the water quality, all of that data also comes into our back end system. If you imagine that a treatment is three to four hours, we collect between two and three million data points per treatment. So we're talking about tens and tens of millions, maybe billions at this point, of data points that we're collecting that we can feed into our systems. So as democratization, it gets me excited, right? Because you start to talk about that data can help support the design team, right? So how are they going to redesign a pump? Or a motor or something to be more reliable. That data might be used to better glean insights into the reliability curve of our fleet over time. How are we doing on our devices fleet as it starts to age? And ultimately, all of that data can then flow back into manufacturing. How can we design, develop, and manufacture a product that has already amazing quality, but how can we make it even better? How can we also start to think around the next gen pumps, sensors, etc., that will help fuel our design curve over time? So, being able to connect that digital thread from the beginning of your manufacturing process, When the part comes into oncoming inspection, through the assembly, finished goods, and ultimately how that component fares in the field, really has created a very streamlined approach to design for manufacturability, for design for reliability, and the other DFXs, if you Liz: will. Amazing. So, Marc, let's talk about one dimension of this new approach to manufacturing that I think is so essential, and that is composability. Okay. The idea that you don't have to build the entire system from top down over, you know, a multi year period. You can kind of build it from the bottom up, take, you know, the first step, generate data, understand what you're learning, then kind of learning by doing and moving to the next step. And that approach to manufacturing, that kind of, it makes it more adaptable. It sort of lets people, um, adjust and learn. It's so important, I think, particularly for the small and medium sized companies. Who don't have a, maybe a clearer sense on ROI, who may not have the capital outlays right from the beginning, and It really helps kind of a learning by doing process. Can you talk about composability and the importance of that to you? Marc: Sure. Only kind of recently did the world of monolithic manufacturing systems kind of see some competition. For decades, and even up until the last five to eight years, there's When you wanted to get a manufacturing execution system, you had to sign up for this 18 to 24 month design and validation period. And the funniest part about it is that the system was designed and built not for the operators who are actually building the product, but for the management that sits typically, you know, behind the glass, if you will. And what did that lead to is that it actually led to issues of efficiency, quality. It wasn't actually designed or intended for the person who's actually going to be using it every day. So, with the entrance of composable MES architecture, it's really providing users, and when I say users, I mean the operators, manufacturing engineers, the process engineers, So, with It's a tool that they can design, customize, and ultimately control. And that is allowing them to run faster, bring better quality into each step of the process, and ultimately allows them to glean the insights from the data that they need to be successful in their day to day jobs. Liz: So what's the impact of this as it relates to vendor selection? Marc: So as you kind of go down the journey of selecting your right MES or right manufacturing execution system, if you want to be a forerunner in your industry, if you want to be a forerunner in your competitive landscape. You also need to make sure that you can maintain that distance after you set up. So a lot of customers, what I find interesting, is when they're looking at their MES, they're thinking about only today. They're only thinking about, how can I get this launched today? And then it's, in five years, it's the next person's problem. But I really think asking the right questions up front will really kind of give you a sense of if this vendor or if this partner, because really they're going to be a partner for a long time journey, is the right one for you. So, are they often providing new and innovative improvements and solutions to the customers? Or are they giving you this one, this is the box you receive, you know, if you want to make a change, it's 10, 20, 50, 100 thousand dollars in the future. Are they able to solve complex problems when they arise? So if you're on the cutting edge of technology, you should expect some level of improvements that are going to be needed and bugs to be found. That's just part of living on the cutting edge. But when those problems come up, are they a vendor or a partner that's going to solve them with you quickly? Do they understand the impact it could have on their business or on your business and want to work with you? And then the last question I always ask is, are they leading the industry or watching others? For example, generative AI comes out. Are these vendors looking and jumping headfirst in, being like, I want to be part of this, I want to provide those tools and opportunities to my clients? Or are they the ones sitting on the back end going, hey, let's see how this plays out? And I'm always one to say that if you want to be a forerunner, you have to have vendors who are also wanting to be there with you. So those are some of the questions that I typically ask when I'm going through the vendor selection process. Liz: Interesting. And then also, what about the whole idea of the So, the creation of a digital fabric, you know, from the device placement through the manufacturing, this idea of a closed loop system, like, how does that work? What are the advantages of that model? Marc: As you were mentioning, Liz, this is the holy grail of IoT connectivity. As devices become more connected. Regardless of if it's in the automotive industry with cars such as Tesla, or the medical device industry with devices such as Tablo, to the Air Force and the national defense industries, being able to correct that digital fabric, being able to connect the true quality of a component and system from how it was assembled, right, even before assembly, how the person was assembled. All the way through to if there was an issue, how that part was returned to the manufacturing facility through root causes analysis. If we can create that digital thread, it allows us to make reliability design, next generation improvement much faster. Because a lot of times, once you ship a device, there ends your ability to develop and improve. Unless you're getting information on complaints or other means, it's really hard to glean the insights of truly what's happening with your device. But if you are able to make that connectivity from how it was manufactured, through field performance, and then ultimately, Back to return of manufactured goods, you're able to create that digital fabric, as you were, you were discussing. And that outset, this has really driven significant improvement in our reliability of our products. Liz: So Marc, let's turn to one of the more important questions here, and that is the role of workers. And how they are central to, I think, the outset story, and also, you know, both you and Natan talk about human centric technology and the ways we augment workers rather than replace them. And that certainly was our finding at MIT when we were doing the task force on the work of the future, that for all of the fear about the robots are coming to replace people, that we haven't necessarily seen that, that we are still in need of workers. Um, but. Perhaps the jobs are changing. Can you talk a little bit about your experience and what it means for maintaining and growing a workforce? Marc: I agree with Natan. I don't think the robots are coming. And I also don't think Gen AI is going to just replace us. I really believe that as we see these technologies, be it robots over the last 20 years or Gen AI over the past few, we're going to see other technologies come out over the next 10, 20, 30 years. And there'll always be that fear that whatever technology comes out, What I really think it means is that we're going to have to train and not steal our talent. That's truly what's going to be necessary in order to maintain a workforce that is effective and efficient at their job. There's kind of a quote that people sometimes say that says, if someone asked me today and said, if you train people and they leave, isn't that a big issue? And something that I've always said, well, wouldn't it be a bigger issue if I don't train them and they stay? And I think that's a big part of what we do at Outset is we really believe that new technology needs to be embraced. Transcribed It shouldn't be pushed aside, and we need to train people on how to use those new technologies. And that we do through what we call Upscaling. So what we do for our operators is we teach them how to work within a digital world. Suddenly we get to the end of the company, and the operator during the interview process will interact with Tulip. They'll be given a touchscreen display with some, let's say like a Lego set of components. And they need to follow the instructions without being trained. We're actually looking to see how do they interact with the digital world. Do they know how to intuitively understand the screens? Do they understand what's being presented in front of them and how the system works? And that helps us both decide and gauge the ability of that individual to Liz: So, Marc, what about how this approach to kind of human centric technology and augmenting the worker, what does it mean for retaining workers and attracting workers? I know right now in the U. S. we have over half a million job openings in manufacturing. It's very tough to attract the next generation and, of course, retention is really important. What's been your experience in that space? Marc: So what else have we seen? Benefit in both sides, both for retaining and attracting talent. So as we think about individuals coming out of school, they all grew up for the last 10 years with one of these in their, in their pocket, a cell phone. And as we see more and more of them are into video games and interactive experiences. And imagine if they're coming out of school, they go to an organization, and everything's archaic, everything's paper based, old fashioned, and there's not really a sense or a desire to innovate. That wouldn't be a place that I would want to work, and that would be kind of a demotivating factor. Well, imagine if they came to a place where they have a digital platform that allows them to create, invent, and design. Applications that are going to make their jobs easier, more efficient and fulfilling, as well as the people who use them on the manufacturing floor. What if we took it one step further, and as we think about our collaborators and our operators on the manufacturing floor, What if you enabled them to actually provide their suggestions in a real time basis through a digital platform on how they think their job can be done better, right? Because ultimately the people on the floor know sometimes what is the best ways to produce a product. They do it every day. How do we kind of streamline that and provide them the opportunity to engage with our engineers and provide feedback? So those are the types of things that we've seen have really helped us both for retaining and attracting critical talent. Liz: You know, that reminds me of a quote we heard from a company when I was doing some MIT research out in the Midwest in which we said, well, what about the lights out factory? Is that what your, you know, ultimate goal is? And the person said, you know, you can't innovate in a lights out factory. All right, you need actually people to help with the innovation and I think that's what is at the core of what you're talking about. Let me turn now to the question of adoption for the small and medium sized firm. You know, small and medium sized manufacturers are the backbone of the U. S. Manufacturing base, 250, 000 of them, right, 98 percent of establishments, over 40 percent of employment. And tech adoption is not necessarily front and center, uh, in terms of priorities and an ability to figure out ROI and, and adopt technology as quickly as large firms. But what you're talking about, and what the sort of composable MES model suggests, is that we've got lower barriers to entry now for that technology. And that, you know, if we can get this in the hands of SMEs more broadly, it really can be transformational for this moment that we have in this country and others around industrial transformation. Can you talk a little bit about, from your perspective, obviously you're the president That's a really hungry startup. But, you know, more broadly, there's just SMEs more generally, kind of, what does that path look like for them? Marc: Liz, you and I have spoken about this in the past and it's near and dear to both of us about the importance of manufacturing in the US, the importance on a strategic level and the need for us to continue to reinvigorate our manufacturing footprint here. And the composable NES architecture is not one just for the big boys and girls club. That is only for the S& P 500 companies. This can really be adopted on a multitude of different scales, also for the mom and pop shops that are looking to have just one or two, let's say for example, tulip stations. If we think about outset, just to give you a level of the scale we're talking about and the speed at which it can be done, we have a product that has 2, 000 components, it took 7, 000 steps to produce one machine. And we were able to digitalize the entire process in under five months. So the time to realization was very fast for a very complicated process. Today, when we want to go and do fixes and make new applications for new products, we can now do it in under one or two weeks, where the barrier is even lower as we become more and more self sufficient. The team at outset is also a team of only four individuals. For in terms of process engineers and manufacturing engineers, supporting two very large scale, uh, manufacturing business units, as well as a spare parts business unit that produces 70 to 100, 000 units in itself a year. So this is all to say that you don't need a large bankroll or a team of 50 or 100 software engineers. You need a few passionate process engineers and manufacturing engineers or technicians. that are excited about new technology, that are willing to try something with one or two licenses. It doesn't take a lot to get going. Liz: Well, Marc, this has been a fascinating conversation and a continuation of the number of fascinating conversations you and I have had over time. Thank you so much for bringing this all to life for individuals, workers, companies, you know, those who are considering the digital platform and a future for advanced manufacturing. Really appreciate it and look forward to working with you going forward. Marc: Thanks, Liz. It's always a pleasure to talk. We're both very passionate about manufacturing, manufacturing in the U. S. And our manufacturing brothers and sisters deserve technology that helps make their job more rewarding, more useful, and impactful. So I'm always happy to talk Announcer: about it. Liz: Great. Thank you. Announcer: Thank you. Thank you for listening to the Augmented Ops podcast from Tulip Interfaces. We hope you found this week's episode informative and inspiring. You can find the show on LinkedIn and YouTube or at tulip. co slash podcast. If you enjoyed this episode, please leave us a rating or review on iTunes, or wherever you listen to your podcasts. Until next time!