speaker-0 (00:05.646)
Welcome to the Hard Tech Podcast.

speaker-1 (00:09.57)
Welcome back to the Hard Tech podcast. I'm your host Deandre Hericus with my usual suspect, Grant Chappell. And everybody, we have a super exciting guest here in the studio, Norbert from Care Medical. Came all the way from Atlanta, so thank you for making the trip. We just had a great tour of Glassport and great conversations ahead of the podcast. And so excited to get this one kicked off and would love for you to just start off with a brief intro on yourself and your background.

speaker-2 (00:14.414)
How's it going, everyone? Thanks for joining.

speaker-0 (00:32.012)
Yeah, thanks you guys and it's great being here at your place. It's my first time in Indianapolis. I drove up from Atlanta. Yeah, thanks for the cold. Appreciate that. So going back to 50s when I'm home. Yeah, Norbert Leinfelder. I'm a PhD in physics. I got my degree back in Austria. That's where the accent comes from. I try not to lose it, so don't challenge me too much on that. But I really have a, you know, my passion is in medical device development.

speaker-2 (00:32.206)
Thanks.

speaker-2 (00:38.761)
You got great weather for it.

speaker-0 (00:59.586)
Going back 2003 when I came to the US, was basically thrown in the cold water as a requirement engineer. Was at Siemens, good big projects, went over to Coliseis Medical, also in the San Francisco Bay Area. And then Fresenius, dialysis, incredible field to work in with amazing colleagues and the clinical staff. I tried startup for myself for a few years doing class three devices.

In San Francisco.

speaker-2 (01:29.704)
You know, challenge. Just starting on hard mode. No challenge.

speaker-0 (01:32.238)
So much easier than class two, right? And I learned so much. I did class one, class two, and class three. So then after that, I went into respiratory care. And that is its own challenge. Starting at InnoGen, basically the market leader at the time where the portable oxygen concentrators were kind of one of a kind. now suddenly you can produce oxygen on the go.

speaker-1 (01:34.644)
Class 1, right?

speaker-2 (01:56.372)
And prior to that, a patient would have carried a tank around of pressurized oxygen, right?

speaker-0 (02:00.686)
So that is the alternative. You have a liquid oxygen tank that's not only pressurized and looks like a bomb, it is also very difficult to refill, very difficult to lug around. So it has its challenges. The industry loves it because of its cost proposition, the last, I don't know, 10, 20 years. So the total cost of ownership is very, very...

speaker-2 (02:16.366)
Yeah, the tanks last almost forever.

speaker-1 (02:21.131)
I think that's what most people think of when they think of oxygen that their grandmother might be getting.

speaker-2 (02:25.464)
They're wheeling around the thing on the cart with the handle.

speaker-0 (02:27.822)
Yeah, I saw a lady at the grocery store. She had a little tank in her cart, hooked up with the cannula. And I asked how she does it. Yeah, I know it's OK. It's a necessity. And then I talked to her about the POCs, the portable oxygen concentrators. And she was blown away by what? can use that. learns the batteries. I don't have to refill. It's lighter. I can use it in a backpack. I can use it in a handbag.

speaker-2 (02:50.87)
Yeah, and it's one of those things that, you know, power is almost everywhere. Going to get those oxygen tanks filled is not everywhere, right? This enables you to travel, you get on an airplane with it.

speaker-0 (03:00.098)
That's exactly right. You cannot fly on an airplane with a tank. With a POC, you can. That's one. Also, the military loves oxygen concentrators. Why? You can maybe lock tanks into the desert. I mean, that's its own challenge, is to fly them around, highly pressurized. But what they do have is power. Military has endless power wherever they set up shop. And so a concentrator runs on power endlessly.

speaker-2 (03:25.42)
Yep, it's whatever you want to fill it with. As long as you have power. Solar generator from a gas generator. However you're getting your power, you can recharge the generator.

speaker-0 (03:32.13)
That's So that's the value proposition here. It's something that you can use in your Ampulatory mode. You can walk around, can go golfing, go hiking, do things that you love. And when you're done, when you're home, you can maybe change to whatever other device or keep being on the POC.

speaker-1 (03:49.434)
I think it's really cool just because one of the reasons we were so excited to have this podcast with you is I do think you're really making people's lives better with the POCs. You've got people that are used to having these large auction tanks and carrying those things around. And now you actually can have something that looks a little bit better. We're going to dive into on the podcast as well how you guys are introducing more industrial design because your customers, they're not getting younger, but they're getting more tech savvy. So I'm excited to dive into that with you. But the first question I've got for you, Norbert, is around what are the core things you've learned about getting

medical devices out the door. You said you've been with everything from class one to class three. What are those core lessons that you've learned about from doing that?

speaker-0 (04:26.316)
Yeah, when you build medical devices, you really have to get them through the FDA. So they teach you the device has to be safe and effective. And those two little words, they make you develop for two to three years. Whereas in a consumer device, can pretty much do what you want unless you kill anyone. It's much easier to get something out.

The rest is business, right? You control the quality through for warranty and service and all this, right? But once you're in a medical space, we always think about the patient. The customer is not always the patient. The customer is not. It's the hospitals, it's the distributors, right? Those are, they buy your products by the 10,000. But it's really the patient that you have to think about. The use cases, how is it being used?

speaker-2 (04:57.964)
It's not. It's often not.

speaker-0 (05:12.032)
What are the error modes? What is the foreseeable misuse?

speaker-2 (05:15.105)
Forseeable misuse is my favorite isn't when you're developing it and you go into your first like prototype user study not even a clinical just getting this in front of users and you know the second person you give it to just turns it upside down and hits all the buttons at the same time and the you know the Person who owns that project why would they do that and the engineers like no let them we need to learn how to prevent this

speaker-0 (05:32.714)
Absolutely. Yeah, we do design verification, you know, by the book, we have sophisticated protocols, we run it with a very high sample size. But of course, you cannot foresee what to do really with that. Right. So ideally, you have a user and that's more in the validation phase, where you have to use a really, you know, mock with it and use it where however they want to be. And maybe, you know, there's the occasional drop, right. So use to all kinds of crazy stuff.

speaker-2 (05:59.31)
One is my favorite, like for people that aren't in med device, break, you know, peel the onion back in this. I like you, start with user needs, how the user expects the device to work. Then we translate that English to English into engineering needs, design input requirements. So those are how long should last on the battery in minutes, not it should last a whole day is what the user expects. All right, how many minutes and hours is that? know, how user expected to survive a drop, engineering, six foot drop on a concrete surface of this hardness. You you document that. And when you verify it, you can verify piece by piece.

The engineer can run a test and make sure his part or his or her part of the design is exactly what their spec was. But you never know if it all works together until you assemble the whole product and give it to a user for validation. Those are your four main steps that most people that aren't a med device don't understand. But the moment you take that process to consumer electronics or one of your other industries, you shortcut all these mistakes you'll make early on in consumer electronics by rushing to the end and then give it to a user.

speaker-0 (06:53.482)
So I agree with you that this is a very important phase to do. The problem is it's also at the end. So you want to learn earlier. So that's why I really believe not only in the rigorous design verification, but do some user studies, do some wear studies all along the way. Use early prototypes. Use the box that you put the device in and do studies with that. The earlier the better. But that's something that not everybody's thinking about.

speaker-2 (07:17.004)
Right, and it's in the medical world, there's so much, I'll call it rigidity around the formal process. Everyone forgets you can do informal processes to go learn and you should. We always joke our alpha prototypes is a shoe box full of parts that generally does the thing. And most people would shy away from showing that to a user. That is the first thing you need to get in front of users' hands. And we'll even like have a 3D printed render or 3D printed sample of what it might look like one day. It will look like this. Ignore the shoe box, man behind the curtain. Just use the product. Does it produce enough oxygen? Does the UI?

on the laptop, know, the touch buttons, this is make you happy. Everyone in med device feels like they have to go to clinical before they start actually soliciting feedback and that's the biggest mistake I've ever seen.

speaker-0 (07:57.71)
I agree with you. That's why I believe also in the feasibility phase. Before you really raise a pen to think of your industrial design or your final electronics and your software, you build breadboards very, very early on where you vet out your technologies. Which components do I use? Do I even have the right compressor? Do I have the right valves? Do I have the right sieve bed materials? You couple that all together and run a bunch of tests with all kinds of parameters that you change in a very safe environment. Nobody's looking.

a safe space and that's time well spent. It's a very small investment compared to when you encounter issues later on in design and especially in design verification. In design verification you don't want to fail. It's not an exercise to say oops we have an error here let's fix it. No it's too late. For me it's a formal exercise so that when you design you already know it works. You already have vetted your power parameters and your reliability.

speaker-1 (08:53.944)
how did you learn that to be important? It's like basically the test before the test. I'm sure that you might have stubbed your toe and you didn't do the test before the test. And so what was that lesson that got you to that sort of pedigree of engineering?

speaker-0 (09:05.486)
So unfortunately the hard way, it's always when you go, I mean you learn from the best, right? So when I started, I followed in the steps of the great ones, great systems engineers I worked with, great mechanical engineers, electrical engineers. They taught me that, here are the risks, right? Here's the risks.

The problem is not so much in the subsystem, right? There's great electric engineers, great mechanical engineers. The software folks are the best and they code the whole night. talked about, know, it's a software developer is a machine that converts coffee into code. those are the people I worked with and I love it, right? The problem is the system thinking. Integration.

speaker-2 (09:47.426)
like integration hell is what we always joke.

speaker-0 (09:50.274)
the system suddenly behaves differently. So in a feasibility phase, you can actually talk about systems, you can put things together and talk to each other. How does mechanical design fit with the electronics of the software? You cannot do this in the design phase where you work on a part maybe months on end, right? You don't see that part until much later when you get maybe a machine part or then a plastic molded part. So that's why this early phase is so important.

speaker-2 (10:17.902)
Well, I think the to double click into that, I think you guys are standing on the shoulders of giants. And when you're going to that feasibility phase today, because you can borrow your requirements and verification validation test plans and documents from the last time, it is so hard when you're starting for a brand new product or product category to write really good verification validation test plans before you've built the device. But it's really important to try. Because if you have those plans laid out, even if they're 25 or 40 percent of what they'll end up being by the end of your first revision.

all the way through the FDA, your engineers and feasibility have a checklist. Because if you give an engineer the ability to prototype and not tell them what you're testing, they're going to make the fun, fast thing. It's not as useful as it could be if you give them the checklist, because then they'll build the fun, fast thing to see if it breaks or not. And that's what's the important, like, connecting the dots here is you have to slow down and eat your veggies and write down what you're going to test before you start building your feasibility prototypes. And the moment you have that, let the engineers lose some dessert. And they love breadboards and prototyping and moving fast, because that's the most

nimble they get to be in their careers.

speaker-0 (11:18.188)
Yeah.

speaker-1 (11:19.08)
Honestly, Grant, curious. asked Norbert, I want to ask you the same question. mean, feel like glass does a very similar process, right? In terms of the medical, like what were the areas you stubbed your toe to get to that point?

speaker-2 (11:29.998)
it's entirely the school of hard knocks. Everyone tells you, you learn from the best and they tell you to eat your veggies and you always end up skipping to a little bit of dessert until you realize you get into the end and you're out of shape and you're like, I should have eaten way more veggies. So thank you to all of our early clients that learned with us. But it's with experience you gain this knowledge of how much veggies to put in your plate to move fast and nimbly and not slow you down, but to prevent yourself from running into those roadblocks later.

speaker-0 (11:55.662)
That's the art of product development. How many veggies and how much meat do you put in your plate before you go to dessert? I love this analogy. Part of it is experience. There's definitely that. You have engineers that have done this many times before. They can point out where the real challenge will be. But the other half is data. So let's use the quality data that we already have. We do have complaints data. We have manufacturing data. We have service...

data in the field, we can talk to customers, we can understand by data how big a problem is, what we want to look out for, and it's okay sometimes to accept a risk. That's very okay, but you have to know which one.

speaker-2 (12:36.494)
Can we dig into that? This is the thing that most people I talk to in product development don't understand is that it's never perfect. When we're putting together timelines at the early stage of a program, I'm never right. It is the best I know today and here are the risks I foresee and there's some amount of unknown unknown risk that even I can't predict at this stage. We have to plan for some contingency here. How do you guys in a more formal process in building similar devices year over year?

How do you guys look at risk and what risks are the typical ones you choose to accept until what stage is the question I want to ask?

speaker-0 (13:08.128)
Yeah, that's very important. So when you look at the medical device development field, the FDA requires you to go under design control. What is design control? Design control is a very systematic approach from the idea to a finished product.

We use very often a phased gate approach, a phased gate process where we start, we talked about feasibility, that's not really where the formal project starts, but right after feasibility, you know what you can use in terms of technology, are we in the cost realm, are we in the size, are we in the weight, right, in the performance realm. Once you have that, you start your formal phase, you kick it off, phase one is where your plans happen and the early risk.

Analysis, we do a product risk analysis. That is how is the product the product as a whole? How is it being used? What can go wrong? How can the patient be harmed? How can be? Somebody else harmed like a caregiver. So we'll how is the machine itself harmed? So we look at very high-level system risks that we Put a requirement against for mitigation. So there you have it. There's another design requirement for you. So all this

All this makes it into design inputs. And this is where you start your next phase. The next phase is all about design. You take what you have learned in your inputs, you do your risk analysis, your requirements, and you go to work. So you basically break down the system into what we call critical components. In respiratory, we have a few critical components in the device. Compressor is one, valves is the other one, battery.

the sieve bed, which is actually actual extraction of oxygen. those critical components, they need to be designed to reliability, to cost, and of course also the performance. So this is what I call the triangle of death. You have the size, you have the weight, you have the cost, you have the performance. reliability, and it doesn't cost anything. That's the challenge we have. So talking more about risk, in this phase, as we develop the

speaker-0 (15:03.234)
components, we also do what we call a DFMEA. So it's a design failure mode error analysis, where we look into each of the elements of the design and ask ourselves, does this actually add risk, or does it mitigate your risks? Is it actually robust enough to foresee those misuses? So that's the DFMEA is doing the design. Then we put things together. So suddenly it becomes a system.

speaker-2 (15:32.974)
you go through integration hell and now you have to talk to each other.

speaker-0 (15:35.086)
And everybody underestimates this. Because the system behaves differently, then the components would otherwise...

speaker-2 (15:42.35)
And it's not linear. If each part of the system took two units of effort, you had mechanics, electronics, and software, and that should be six units of effort, when you combine them, it's always non-linear. It's a multiple of that, and it's so hard to predict. Is that an experience that you also...

speaker-0 (15:55.936)
It always takes longer than you actually think it would. And so that's where the experience comes in. that's where, you know, I really believe in good systems engineers. But system engineers don't grow on trees. It's very hard to find them out of college because there's not such a degree necessarily. You have good biomedical engineers. They're the closest to what I think of a systems engineer.

But then you have your other design engineers, mechanical, they very often have system thinking. And that system thinking is so essential when you build a device. Because I'm not only looking at my mechanical enclosure or my brackets or my doors, exactly, and all the little filter claps and all that. No, I need to think about that thing runs, right? It creates heat. Exactly.

speaker-2 (16:32.2)
Your battery door.

speaker-2 (16:42.53)
How am gonna get out of this box?

speaker-0 (16:45.902)
And nobody else thinks about it. The project manager doesn't know it, it says 1 plus 1 plus 1 is 3. What's so hard? No, it is not. The system has its own dynamic. And that's the key.

speaker-2 (16:56.724)
I love that and I think that in that dynamic, the multiplier on the non-linearity of the challenge of system integration gets bigger at the later in development. Right, you know, if you need to change a connector on a board, this location, it's really easy to do before that board's been revved yet. It's really easy to do before the mechanics have been designed yet. The moment you're already integrating and we need to move a connector on a board, the mechanics have to update, the board has to update. Heaven forbid, software has to update to detect that, you where the connector moved to or things like that and it's this, everyone now has to touch a

change even if only was driven by one of the disciplines. So it's the later you wait to do systems integration the more nonlinear any change becomes.

speaker-0 (17:35.028)
I love the example. Move a connector on the board and your thermal flow changes. So now suddenly you have two degrees more in your device. no, I need those two degrees because I'm already at limit of, you know, when... Yeah, compressor and so yeah, very, very tricky. So simulations are important, right? You don't want to find out after you put everything together. Do as much as you can in the computer. That's why we have these great software packages.

speaker-2 (17:45.848)
processor, yeah, with a compressor.

speaker-1 (17:58.542)
Nor, you guys have got this new element you're introducing to care medical, which is industrial design. And so we chatted a little bit, like I mentioned before, when we were first having the conversation on how industrial design is being introduced to the medical world.

speaker-2 (18:12.107)
one more team to integrate.

speaker-1 (18:13.358)
Which is one more team to integrate. It's a new skill set because consumers, like you guys mentioned, you guys are focusing on the consumer. Though they might not be your customer, that's the person that ultimately is going to be using that product. And so what have you guys seen? You mentioned people become more tech savvy. They want to look good with it. And how are you turning this life-critical medical device into something people want to live with?

speaker-0 (18:35.33)
Wow, industrial design is a field that spans everything. It spans the mechanical world, spans the usability work. Even from a business perspective, this is where your brand language lives in. So it is very important, at least I found out, it's very important to do that early on. Don't treat industrial design as an afterthought. Don't be rapper, I designed the device, let's do some industrial design on it.

speaker-2 (18:59.244)
It's not a wrapper at the end of it.

speaker-0 (19:05.11)
What I believe in is have those industrial designers in come early, talk to the marketing folks. What does your branding look like? What is the message that you convey when you look at such a device, when you hold it, when you feel it? And that's one thing. It's not that we choked about.

A medical device is basically a box that's beige, that has an off and on switch, and everybody's happy. So does the patient care? Well, I tell you, the patient landscape is also changing. The patients, they become more tech savvy, unfortunately become younger. Everybody has iPhones, they want to look cool. They don't want to have a medical device next to them. It shouldn't look like a medical device. So how do you do that? I would.

challenge that one of our mechanical engineers or system engineers actually have the answer. They are asked to build the smallest device possible. What do they do? They put a box around it.

speaker-2 (20:00.942)
The smallest box they can shrink wrap around the electronics, compressor.

speaker-0 (20:05.038)
And the industrial designers, they come from a different realm. They say, yes, it needs to be small, but it also sits on your body. How does it actually interact with your body? Is there a curvature involved? Is there some shape? Is there a way you can actually operate it better?

speaker-2 (20:23.704)
Where should the buttons be? If you wear it as a fanny pack or on your hip or as a strap, and can we put the buttons in a place where all three use cases work?

speaker-1 (20:31.246)
And it might be counterintuitive, right? Because you might think, OK, they wanted to be on their body. It needs to be super small. It's like the small box. Well, I mean, there's tons of people these days wearing kind of the fanny packs that go around your shoulder or go around your waist. It's like maybe it doesn't have to be small, but it has to look like something else that you would wear otherwise as a fashion piece,

speaker-0 (20:47.788)
I agree. we all want, you know, in oxygen devices, what do we want? The performance is measured by liters per minute of oxygen. The higher the output, the more oxygen you get. So your population of patients grow, the more of these output settings you can offer. You have one device for a patient that just got diagnosed.

speaker-2 (21:04.75)
Right.

speaker-0 (21:08.6)
with COPD versus somebody that's unfortunately two or three years living with the disease needs more flow. How do you bring this into one device? You do more settings. More settings means you need a larger compressor. What does a larger compressor do for you? Size, battery, battery runtime, cost. So that's the triangle of death where we're operating. An industrial design needs to rein this in.

speaker-2 (21:22.094)
Bigger battery.

speaker-2 (21:33.57)
They help put a lens in which to see perspective on the triangle. I love your triangle of death analogy. We have one called the triangle of fun. Good, fast, cheap. Pick one or two. And then the third one just completely is uncontrolled, right? Yes.

speaker-0 (21:45.614)
Do this in a barbershop. Good, or cheap, pick two.

speaker-2 (21:48.886)
Yep, no, I love it. But that's, I think it's so important that what you just said, that perspective that industrial designers can give an engineering team. Engineers can almost meet any goal. They're very clever. They'll solve problems incredibly well.

They're bad at picking what goals to meet. That's their laser focused on solving the problem, not coming up with the right one. And industrial designer UX UI engineers are so good at asking the right questions and putting themselves in the user's shoes. It's that empathy thing that you and I were talking about when we were interviewing engineers at college. They look at your shoes, my shoes, or me in the eyes. Who are we talking to here? Yeah, I love story. Yeah, Norbert and I were talking about how do we recruit engineers? Because you're also in Atlanta, which is like middle of America, right?

speaker-1 (22:18.914)
Grant, can you share that story live with us?

speaker-2 (22:27.5)
We're not Boston or LA. We're in Indianapolis. We're in the middle of America. And how do you recruit these really talented mechanical electrical systems engineers if you and I can both find them and fight over them or industrial designers and have them, you know, attracted to these smaller cities that aren't on the coast? And we always said, you know, we're interviewing at local colleges and how do you make a snap decision if this is the right person for your team? We always joke if they're looking at their own shoes, they're really introverted and they're going to be a good principal engineer one day. They're going to solve math problems. If they're looking at my shoes, they're a good balance and product development and systems engineering.

And if they're looking at me in the eyes, business development sales, let's go talk to people. And there's these three ways you can figure out who you're talking to very quickly just based on eye contact. And I'm joking for the record for everyone listening. This is good narrative to do a quick filter.

speaker-0 (23:09.298)
I love it and and the working with engineers is so much fun. It is and if all these different types No, nobody is nobody's in the same. It's they all they they think differently. They they have different values What what unites all of us is the purpose, right? So that's probably where they came to medical device development in the first place or to health care is What are you building this for? You you extend the life of a person of a real person. It could be your relative one day It could be yourself one day and we have some examples

where such of those patients they became actually engineers to work in that. And children of parents that live with the disease and they are now so proud that they do something meaningful and build those devices for future generations to come.

speaker-1 (23:54.6)
I think it's great, honestly. And getting into the world, it's like teachers, for example. They get into the world to teach. And my girlfriend does jokingly say she got in the world teaching to change lives. Now she does it for the vacation. But I'm sure that's not the case for the engineers. Grant, I got a question for you on Glassboard historically had some people that would dabble in industrial design. But recently, I feel like the firm's taken a

Stronger approach and like actually bringing on new the industrial design team and building that out What have you seen leading, you know, historically I just a full team of mechanical electro, etc Introducing industrial design. What have you guys had to work through to kind of merge those two? Like do they? Do they design at the same time? Yeah, I'm curious about how that integration really looks downstream

speaker-2 (24:34.476)
play nice together.

So it's interesting. So Glassboard started as a small engineering services firm in my mom's basement with Drew and I. And we had very limited skills at the time, but we grew ourselves year over year. We grew the team and hired new skills year over year. And we've always been expanding in new skills, both because we want to and most of time because our clients ask us to do one more thing. Industrial design has always been one of those. But just like regulatory or quality, which we don't yet currently offer in-house, but maybe one day, we've always had wonderful partners, right? There's a lot of industrial design consultancies or firms, or I know you're working with some right now,

that exists and they only offer industrial design and they're really good at doing it as a third party, empowering the end client to run with their dream and not hold up the process or partner like with us as Glassport, an engineering firm that would come in and pick up their design language and carry it forward. So we operated in that manner for many, many years and we started hiring some mechanical engineers that had industrial design tendencies. Then we hired a dual major that had both degrees and now we hired our first full industrial designer and built a team that is now dedicated and separate from our mechanical team.

And we are doing the same work we always partnered with firms to do. We're still partnering with firms. We're happy to work alongside people. We're never picky about who does what. It takes a village, as they say. But the one benefit we've seen internally is at Glassboard, we knife fight all of the time about design decisions. We throw down in the conference room about the product, the user, why it should be this way. everyone here cares and gets emotional about it, right? They're putting up in arms why this feature should exist.

And it's so much easier for us to do it within our own culture because we're pretty abrasive. And you've seen this. We're passionate and we will knife fight about our passion for why this product should exist or how it should save lives this way or that way. And doing that all in the same conference room just shortcuts that process. Being able to, you know, the industrial designers say, hey, it needs to be this way for usability. And the ME say, but that's going to decrease battery life because it won't be as big. And then figuring out what's the optimal solution. That's been awesome. And the flip side is at the end.

speaker-2 (26:32.566)
Once you go through all your design and all your design for manufacturing and your mold vendors had input on all your draft angles, the ID team is still there to judge those final tweaks into the vision they set out for within real manufacturing product requirements. So having it in the house has been beautiful and beneficial, but third parties are amazing to work with because they bring a different perspective or they might be an industry expert in a certain industry like cycling or med device and they can bring this bespoke industry expert for that team and that industry.

in one place where we're generalists. So I think that's the pro con about it, but we've really enjoyed having it in there through the creative. So they're always throwing wacky ideas out there and mechanical engineers have to go full, you know, figure out a pull it rabbit out of a hat doing.

speaker-0 (27:14.54)
Yeah, and at Care, we don't have industrial design on staff today. It might change in the future. So as you said, we partner with somebody that can do that. And I agree. It is not an event. It is a process. They live with the engineers throughout.

Of course, the beginning is very important because this is where you learn about what are the requirements, who are users, how is it used, where is it used, when is it used, what is your, we talked about the brand language, right, that gets transferred into design language then. But they need to stay throughout. So that's the plan. And I love the skills of those guys, right? You said right brain, left brain, right? Equally present, so hard to do.

And yeah, let's see what comes out. I'm very excited.

speaker-1 (28:03.382)
Yeah, one of the questions I've got for you is whenever you introduce industrial design, typically you're starting to talk about like miniaturization and how have you guys worked on like miniaturization while maintaining reliability with products like this? like what's been that process? are like kind of the difficult areas of doing that?

speaker-0 (28:19.886)
That's a great question because it's a very difficult question. Everybody expects us to build the lightest, smallest devices that have incredible battery runtime and lasts at least eight years. That's where the industry goes to. And by the way, it shouldn't cost much because cost pressure is there simply because competition woke up since COVID. there's a lot of those manufacturers flooding the market with cheaper products.

speaker-2 (28:28.992)
And reliability and lasts forever.

speaker-0 (28:46.254)
They claim the reliability, but nobody has proven that yet. It's not ages yet. So we have to design to that. We have to bake this into the requirements. And when it comes to miniaturization, there's a little bit of an art here. So you have to package your components. Packaging, I call it packaging. It's the art of mechanical design where you structure the system in a way that your components are balanced in the system. So you have equal weight distribution.

speaker-2 (28:49.902)
The clock is still running.

speaker-0 (29:15.074)
very often the compressor is at bottom because it's the heaviest piece alongside with the battery. And then you have your chemical, your absorption material cans somewhere on the side or somewhere where you can replace them quickly. And then you have to place the electronics board somewhere. Now, where do do that? Well, I don't know. It depends on the size of the board. So...

There is this integration hell, as you say. It's easier to qualify the components that work for you, but then to package them together. But you also create airflow. You need to keep the device cool inside. Nobody thinks about it. Yeah, switch it on. It makes a little buzzing sound. Oxygen comes out. What's the problem? Because the oxygen that comes out is actually room temperature. That's great. There's no surprise there. But what's happening inside? Now, you cannot go up to like 100.

100 C in there right right your plastics will start cooking And it exactly right so what is the number one recall reason in the in the field? It's fire and And what so just smoke alone is a reason for recall is a reason for the FDA to get very worried Oh, yes, and and we know when electronics get hot they smell

speaker-2 (30:07.854)
You're smelling funny.

speaker-2 (30:17.472)
interesting.

speaker-2 (30:27.256)
They do and it's really hard to put the magic smoke back in the electronics. We were it's easy to get it out We haven't figured a way to put them back in yet

speaker-0 (30:33.09)
That's right. And it's actually funny because sometimes we encounter components that have some sort of coding on it. And they start to off gas even more so when they're operating hard. And so that's what we're fighting because you don't see that at the beginning.

speaker-2 (30:38.05)
Like an off-gassing.

speaker-1 (30:38.926)
Yeah.

speaker-1 (30:46.87)
Especially like an oxygen system if if a patient sees any smoke at all they're like I'm breathing this

speaker-0 (30:51.68)
This is good.

speaker-2 (30:54.38)
And I think in the miniaturization game, you guys probably have a blend of off-the-shelf components from suppliers that you're integrating and custom components that you're designing. As you start to shrink that size envelope, you're probably having to go more custom than off-the-shelf as time goes on. And what has been the, you know, what are the big, what I'm going to inflection points? I'm assuming compressor and battery are probably two of the big ones out of the gate that, you know, are your size and volume drivers.

speaker-0 (31:20.446)
Customization, yeah, even compressors, you can buy them off the shelf, but then you basically you... You're just up with the competition, so you don't differentiate. So yes, customization is key there. We want more performance, With lower power consumption, lower heat. Sound is very important. Needs to be quiet. How do you do that? Well, if you drive up your air intake and your, you know, displacement,

speaker-2 (31:28.054)
You just want the guts.

speaker-0 (31:49.004)
It's like a V8, it's louder. And also the perception of sound changes. We're very much about the perception. It's not the absolute dBs that you measure, it's also about how does it sound.

speaker-1 (31:59.912)
Is it high pitch? it low tone?

speaker-0 (32:01.624)
The high is low, is it vibrating, it rattling?

speaker-2 (32:05.004)
Is it buzzing or rattling or is it just humming? Right, it's all resonance. System integration hell.

speaker-0 (32:09.71)
So how do you verify for that? Exactly. So there is a lot of validation happening there. So that's where customization is really needed because you need to, the compressors are typically on a suspension. There's typically some springs involved and some saw. Every manufacturer has its own method. And then you customize every little plastic piece, every little bracket that you buy. I think the only pieces we do not customize are screws. And even those are sometimes not.

what you can buy at Best Buy. So yeah, that's the challenge. it's not just an integration exercise. It's also how do I design those parts to begin with and the materials I choose. And they all need to be biocompatible.

speaker-2 (32:40.362)
McMaster, yeah.

speaker-2 (32:55.752)
and cost because the last thing is once you start customizing things that were off the shelf, your costs just change overnight because you've dropped volume. Right? It's just your volume, not you and five of your best friends splitting the volume of this one key component manufacturer in your industry.

speaker-0 (33:11.158)
Yeah, and that's the story of this industry, really. You want lower costs so that the DME's of the world can still profit from the reimbursement they get. So there is this game of your own, what we call the total cost of ownership versus what they actually can make back through reimbursement or through the insurances.

speaker-2 (33:29.326)
Well, I have a question on that. as we started talking about, hey, let's start designing these products more for beauty and UX and user experience. Do you see a market stratification? Is there a space at the top of your market for Cost Plus, right, where it's the reimbursement, but the users are writing cash to get the smaller version on top of their insurance costs?

speaker-0 (33:50.336)
There is. So the two business models are the DMEs that actually they buy the devices in bulk and then rent the devices out for as long as the prescription goes. they want the device to last for a long time. So the differentiation there you can do is reliability.

speaker-2 (34:05.822)
that's where you can get more cost for them.

speaker-0 (34:07.938)
lower cost of ownership so you don't have to service them as often, you don't have to repair them as often. So that's one. The other big channel is direct to consumer. This is where you can actually have a higher margin because consumers like when I look among ourselves, we are probably willing to shell out a little bit more, buy the device outright because it's for us, it's for our health, it's for us, it's for our social presence out there. This is a really cool device and it can do so much. It costs a little more but I'm

very happy to do that because I will have that for a few years and it is for my health. And you use it every day. So, but this is the same process when you buy a phone, right? You have an iPhone, do I need this or do I need that? Well, and then you start thinking, I have this probably for three, four years. Is it a dollar a day? Is it a dollar 10 per day? Well, so here's the decisions being made, right? And we see that, we see that with younger customers, customers that are...

speaker-1 (34:41.429)
And I use it every day.

speaker-2 (34:52.61)
Yeah, is it a dollar a day?

speaker-0 (35:03.564)
customers that are patients, but also customers that maybe dabble in the space. They don't want a device that can do everything. They want a device that's just right for this prescription. It's probably just one flow setting or a couple. That means that in two years, if your disease progresses, you probably need another one.

speaker-2 (35:21.602)
You get to make two sales per customer. That's right. But they get to save money up front in case they don't need to progress. I mean, this is the ultimate model of do you buy the thing that's adjustable or do you keep buying sizes of clothes as you grow? Right. That's the ultimate challenge.

speaker-1 (35:35.278)
One of the last questions I got for you, Norbert, is just, think you've already kind of alluded to it and hinted at it. I mean, what does the future of medical device, either design or functionality, or how we actually interact with those devices, whether it be direct to consumer, which I feel like I don't remember that being very common. It seems like that would be a brand new thing to the industry. And you guys also have this industrial component. So what's kind of the future medical device is in general? In the future, how do you feel like we're going to be interacting with them? And then two, what's like the future of care and like your guys' direction on which

you guys really gonna be focusing on.

speaker-0 (36:06.742)
Yeah, so one thing we haven't talked about is the digital health. It's the connectivity where you have your devices connected to a digital ecosystem that you may own or somebody else knows. But the bottom line is that at the end of the day, you have two problems to solve. You want to know how is your device performing. You don't want to find out how your device is performing through a service call, through a return, through a repair process. Incredibly expensive. Frustration on both sides.

that kind of drags on your performance. But also you want to understand is the patient compliant? Is the patient actually using the device as indicated so that we can make sure the device is right, that you actually have a positive outcome, and you can measure that?

speaker-2 (36:54.606)
and you can associate the correct use with outcomes. And this is what I think everyone misses and what you're just talking about, this IOTification of medical device. Everyone tracks their device. I think the customers or partners I work with that are really doing well in the space are finding a way to track outcomes for these patients in the same data stream. Whether that's through an electronic health record or their own software applications, either or on-device feedback, where the users are giving feedback on-device. If you can tie outcomes to use,

Now machine learning AI gets really fun. And that's what you and I were talking about here before we got on the recording here. And I think for you guys, as we were talking, the maintenance side is the low hanging fruit, right? This is the obvious reason for you guys, because you have a consumable in the device that's not just energy. have these, again, zooming all the way out. The way you pull oxygen out of the air is you compress it and put it through a call a chemical filter to separate the oxygen out. Is that about right for the listeners? And those chemical filters last a very long time.

speaker-0 (37:49.87)
That's right, yes.

speaker-2 (37:53.41)
but they are replaceable just like your RO filters and you're maybe drinking water under your sink. Is that about the right analogy? That's a great analogy.

speaker-0 (37:59.532)
Yes. You have to replace them. so ideally, you replace them a little bit before they start failing and your oxygen output goes down. Because that's the worst possible outcome. When you need oxygen, you need oxygen. You cannot just say, it's not good. I will replace it in a couple of days. So you want the device actually know about it preemptively and maybe even phone home so that your service organization can proactively

speaker-2 (38:18.115)
and preemptively.

speaker-2 (38:25.942)
L-U-F. L-U-F cartridge. Exactly. I love that.

speaker-0 (38:28.216)
So that's just one use case. There are many. The last thing you want are user alarms. Every alarm, there's alarms for all kinds of reasons, for battery overheat, for...

speaker-2 (38:40.024)
battery low.

speaker-0 (38:40.866)
Low, yeah, battery low, low oxygen, whatever, right? All kinds of failures. It's like in a car, but you have this engine light. You don't want the engine light to come on when you are on a mountain because you're hiking. So you don't want that. So we want connectivity to know about the device. And the DMAs, want fleet management anyway. Where is my device? What is it doing? Does it have any problems?

speaker-2 (39:04.236)
Jim's prescription ran out, I need to go collect that device, rent it out to Bob, let's go connect these dots, right? So then again, this is all maintenance and fleet management. So you're using IoT for an incredibly well trodden path, no blue ocean risk, right? We went and talked about risk earlier. You can go copy other industries that have used IoT for fleet management and maintenance and really learn instead of the shoulders of giants. Well, simultaneously, if you set yourself up correctly, collect data now for your future.

learning machine learning about prescriptions about how much flow rate people actually need in their treatment in course of their disease. Because what you might find is for certain people, you can get away with a lower flow rate or less often use and shrink the device and the battery, but you'll have to have data to justify that claim to be the market leader then.

speaker-0 (39:49.39)
That's exactly right and that's where the magic comes in. The data will be the holy grail in the future of healthcare.

speaker-1 (39:55.15)
I'd imagine. I even think that from a business perspective, you think about you guys are selling through the DME providers, Rotech being a local one here in Indianapolis that primarily sell the old-fashioned auction tanks. And you guys bring that data to them. Probably one of the reasons they push that, like you said, is because they last the longest, they're most reliable, most predictable. But if you guys can actually introduce that predictability to your product now, you can potentially start to replace those DME sales reps who be pushing that product as opposed to the other. It sounds like that's kind of a business opportunity for you guys.

speaker-0 (40:24.736)
It is, and you brought up liquid oxygen. We do have products that are amazing machines, but they're highly mechanical. There's no electronics in it, so how do you know how they're doing? So we kind of dabbled in that space a little bit. We're expanding this performance of those machines. They actually now have an IoT solution that tells the DME about the fill level of the tank. So they can come out and refill in time before you actually create a service event.

speaker-2 (40:46.732)
of the tank.

speaker-2 (40:52.886)
Very cool.

speaker-0 (40:53.898)
We developed that and that is something that's very proud of.

speaker-2 (40:56.664)
Can I ask why would a user today be on liquid oxygen? Is it because it's such a high flow rate that concentrators can't keep up? Like why would a user choose liquid oxygen or prescribe that?

speaker-0 (41:06.496)
Yes, it's primarily the flow rate. And one of the big advantages those tanks have, they're absolutely silent. So if you need high flow and you can sleep next to it, you don't hear anything. So there's no compressor running, like there's a fridge next to your pillow. So that's still something where the concentrators are lagging behind. The comfort.

So that's, would say, and in Europe, the liquid oxygen tanks are more prevalent. They are much more than the concentrators. The US has begun switching over to concentrators because it's high tech, it's easier to use, it's more predictable.

speaker-2 (41:41.678)
And we have more geographic area, usually. We live farther away from our service providers, so getting those oxygen tanks is more distance. The are different. Yeah, logistics are hard here.

speaker-0 (41:49.346)
very different in the US. The sheer amount of miles and roads to fill, yeah, that's one of the reasons.

speaker-1 (41:56.706)
Well, I'll tell what, I'm kind of looking forward to getting old now, I guess. I know that I'll be able to breathe pretty well and the device I'm using will be good as well. Thank you so much for coming up to Indianapolis and doing the podcast with us today, Norbert. It's been absolutely fantastic. Everybody, this is the Hard Tech Podcast. I'm your host, Deandre Hericus, my co-host Grant Chapman.

speaker-2 (42:15.448)
Thanks for joining everyone. has been a great one. Norbert, thank you so much for coming on today. This has been a blast.

speaker-0 (42:19.95)
you guys I really enjoyed it

speaker-1 (42:21.236)
Awesome, we'll see you guys next week.