Lab to Market Leadership with Chris Reichhelm

Graphene Commercialisation: Scaling Advanced Materials | Tony Pearce

• Deep Tech Leaders • Season 1 • Episode 6

Share episode

Dive into this insightful episode where Chris Reichhelm interviews Tony Pearce, COO of Paragraf, to discuss the journey of bringing groundbreaking innovations in the advanced materials sector from the lab to the market. Tony explores the critical aspects of scaling up graphene technology, focusing on the importance of early-stage market credibility, manufacturing repeatability, and the alignment of product development with customer needs.

Listeners will gain valuable insights into how Paragraf has navigated the challenges of transitioning from research to production, particularly in the context of magnetic sensing and battery manufacturing applications. Tony shares practical advice on the dynamics between commercial, technological, and scale-up teams, emphasising the necessity of a cohesive approach to ensure successful product commercialisation.

Throughout the conversation, Tony also reflects on his experiences and lessons learned, highlighting the significance of flexibility in manufacturing processes and the value of building strong customer relationships. This episode is a must-listen for anyone interested in the commercialisation of advanced materials and offers valuable guidance for startup founders, tech entrepreneurs, and other industry professionals.



Let us know what you think...

Learn more about Lab to Market Leadership: https://www.deeptechleaders.com

Follow us on LinkedIn: https://www.linkedin.com/company/deeptechleaders

Podcast Production: Beauxhaus


Tony Pearce:

Be prepared to, to have to develop a range of core product to help your customer implement your technology. Don't expect them to do all the legwork for you. Help them to get your, your product to market because, uh, you know, you're on that journey with them and you share a mutual benefit in, in being successful. Don't just expect them to work it out on their own

Chris Reichhelm:

Welcome to the Lab to Market Leadership podcast. Too many advanced science and engineering companies fail to deliver their innovations from the lab to the market. We're on a mission to change that. My name is Chris Reichhelm, and I'm the founder and CEO of Deep Tech Leaders. Each week we speak with some of the world's leading entrepreneurs, investors, corporates, and policy makers about what it takes to succeed on the lab to market journey. Join us. When do you start thinking? What are you thinking about engineering and scale up? And when is it the right time to bring on board a scale up and engineering lead? That's what this week's podcast is about. My guest is Tony Pearce, one of the UK's leading semiconductor and materials, scale up and engineering executives. And he's going to talk through with us his current time at Paragraf, where he is COO, responsible for their scale up engineering and production. He's going to talk through the early days, going to hopefully talk through the, the dynamic of the early team, how they settled on commercializable and manufacturable routes to progress and talk through the general challenges and milestones that have to be considered as you take a materials or semiconductor company from lab to market. I hope you enjoy the podcast. Tony, thank you so much for joining me today.

Tony Pearce:

Thanks for inviting me in.

Chris Reichhelm:

I am delighted that we're getting the chance to have this discussion. Um, I, I want to jump right into it. Let's say I'm an entrepreneur and I'm setting up a company that, I'm setting up a young materials, or a semiconductor company. I've got some ideas and I've been working on these ideas for, let's say, five years, maybe even 10 years, and I've got my ideas now and I want to incorporate a company and I'm going to raise a little bit of finance. At what point should I start thinking about bringing in someone who can help me consider the scale up and engineering aspects of what I'm doing?

Tony Pearce:

Well, you can expect me to say straight away, Chris. COO is a great value for money. You should, uh, hire them, hire them right at the beginning. This is probably where, um, sort of hard tech businesses differ from Software or services or something like that, where you, you're physically manufacturing something, you need to put together the infrastructure and the processes to get into manufacturing something yourself within your own four walls. And that is often a, uh, a step which for somebody that's been working in academia or research environment for a long time, It's both, both daunting, but most significantly it's, it's different and you have different considerations. So when you're researching something, most of these great discoveries are born out of a sort of controlled chaos, if you like, there's this, this, this, what you want to happen in a lab is variability. You thrive on that. You want to be looking at how various material properties in our case, react to changes in conditions in manufacturing. You want the complete opposite. You want everything the same. You need to be repeatability. You don't want a sort of a wheel of fortune effect on the output that you don't know what you're going to get until you've until you've done it 20 times to get 20 different results which which is great if you're investigating that variability and understanding the process, but it's no good for for manufacturing. So having somebody on the, uh, you know, within that group relatively early on, not to burden the organization necessarily with big corporate operational overhead, but to To think about how are we going to manufacture this product further down the line. What is the sensible way to, to, to move forward? I mean

Chris Reichhelm:

And at that stage though, you might not even have a product. You just might have, so if you're thinking, you know, if we're thinking about the very early stages here, we've got a material or we have a process. We have a process for depositing a particular material on various substrates and increasing, uh, their, their performance quite dramatically. And so if we, you know, this works at a particular scale at maybe the gram scale in a lab. But there's nothing to say it's going to work when we try to scale that. There's nothing to say that the unit economics are going to work, or we're going to have a bill of materials that's remotely possible to obtain for, again, you know, within a set of unit economics or for geopolitical reasons, or, or for other things to do with supply chain or their potential quality performance issues. Whatever it is, there are all sorts of different things that could go wrong. So is there a. Is there a way of, you know, balancing out some of that early, wonderful, innovative research with some reality too? And what does that dynamic have to be like?

Tony Pearce:

Yeah. Yeah. I mean, you have to do that. If you think in Paragraf's case with, with Graphene, I mean, Paragraf's whole existence as a, as a company is on the back of this very issue, is that you think Graphene was discovered 20 years ago now. And when, uh, Professors Geim and Novoselov um, did that discovery in, in the University of Manchester famously with a, with a piece of sticky tape and a block of graphite. The usual experimental apparatus. Um, you know, it was a great discovery. It earned them the Nobel Prize, um, further down the line for the significance of that. I mean, there's no Nobel Prize for manufacturing. And therefore, the See, that incredible discovery unlocked a whole load of research into that area, into that material system, into the potential benefits it brings to a whole, whole host of areas, uh, and in our case, the most interested in electronics. But there's, there appeared to be sort of a path of, of discovering, uh, how the material can be used at prototype stage for making great discoveries to write papers with, but not necessarily the one that the industry struggled with is how to manufacture it. And Paragraf hit that, that issue head on, rather than looking, um, at the product first, it was born out of, um, needing to find a manufacturable process. So in a way, that, that issue was, was right at the heart of, of, of Paragraf's creation is to, is to look at how to take a scientific discovery from, uh, the academic realm and into manufacturing. You were in there

Chris Reichhelm:

quite early, weren't you?

Tony Pearce:

Yeah, the company was a year and a half, uh, from spinout, um, and, uh, the, the team when I joined did a fantastic job of bringing that science out of the, the lab environment, um, and into the first premises. And, and, you know, when I walked in the door there the first time, um, it was incredibly impressive to see, uh, at that stage on a, on a modest scale in the, in the early days. What was sort of inside the, behind the door in the, in the, in the first facility that had been done on a modest budget. Yeah, so there's a spinout. Um, but it was already beginning.

Chris Reichhelm:

Where were they in terms of technology progress?

Tony Pearce:

Uh, from a, from the process, the, the being able to create, you know, the core IP had moved on really well. What was it was at the early stages was the productization if that's the best way to describe it and there was the beginnings of the first demonstrator product. Mm hmm, which I think at that stage one two years in we probably believed it was closer to market readiness than, than it really was because you, you learn as you go on and you look back at it, you know, we, there was a lot we didn't know.

Chris Reichhelm:

Yeah. Um, so would this have been like a TRL five or was this a TRL six?

Tony Pearce:

Yeah,

Chris Reichhelm:

if you had to put it into the

Tony Pearce:

Yeah, yeah. It's from a, from a product point of view, yeah. Really. Five.

Chris Reichhelm:

Yeah. Yeah.

Tony Pearce:

Um. You know, and I think we had, uh, we knew the, um, and saw the properties of how graphene can transform the performance of a device. Um, and in those early days, it was trying to get the repeatability to make sure that we could both improve material quality, but do it every time. Um, and doing, and, and, uh, many of the sort of leadership team in Paragraf are coming from that. So we can look at the background, not just myself.

Chris Reichhelm:

Yes,

Tony Pearce:

and that means we kind of, we understand what the industry is looking for and what it needs and therefore we're looking at it from that angle rather than trying to publish headline results that would ultimately, you know, lead to either unrealistic expectations or issues further down the line.

Chris Reichhelm:

You've got a, you have a tremendous amount of experience in, in scaling up semiconductor companies and processes and materials and so on. Is there a, is there a particular dynamic that you think needs to be in place at the very early stages between, let's say, the product or market focused part of a business, the technology, Part of the business and then the scale up and engineering part of the business. This is the way as I kind of think

Tony Pearce:

of

Chris Reichhelm:

it, the commercial angle, the technology angle, and the scale up angle. Is there a dynamic that has to be there? You think,

Tony Pearce:

you know, there is, and there is, I mean, that actually is a part that I think I probably learned the most from my time with Paragraf because I've been lucky enough to be part of a number of situations in larger companies where we've gone through some very aggressive, um, scale ups and ramp ups, but within organizations that are already a little bit more established or have been through a previous round before with a different product. And, um, and in that case, much of your, um, the, the relationships within the different departments, the infrastructure, the internal workings of the businesses is already in place. But when you're, when you're doing that in a, in a startup environment, um, You, you need to be incredibly well aligned at the individual level, not from the commercial customer facing and the research lead, um, and manufacturing operations. And in our case, you know, we sat around in the same room and the same table and we were part of each other's conversations every day. And, um, at the earliest parts, while COVID got in the way of that a little bit, in terms of our ability to all be in the same room all of the time, we, the importance of that alignment was, was really important. You don't have that at the beginning, then, you know, what happens? Commercial will be trying to go down a route of commercializing something we can't make, technology will be making things we can't sell, and manufacturing will be making the easiest thing possible, right? So you have to put, that triangle needs to be somewhere in the middle that, you know, we are aligned all the way through and we understand what the customer base needs and are not wasting time.

Chris Reichhelm:

Do you think that demands of everyone? A finer appreciation of first principles, you know, the, you know, the fundamentals, that kind of underpinned in the technology sense, in order to be able to start scaling something from the smallest piece, if you like, the smallest stage. The smallest scale. Do we need more of an understanding of kind of first principles? Then, for example, you need when you've already done it, let's say, uh, at the ton scale, and now you need to go to multi ton scale.

Tony Pearce:

There is. I mean, we, we, some of the principles in the way that we've gone about going through those decision making processes or making sure that we. Um, considered all of the various aspects, I don't get caught out by surprise when we get sort of close to, to prototype, is to, is to actually put in place, uh, some structure around new product introduction and, and a lot of companies have NPI, um, established NPI processes and, and they can be quite an overhead in certainly a larger organization. Um, but by stripping that down to quite a, um, a light set of, of requirements for the You might call them product readiness levels, I guess, or going from, from initial sort of, uh, commercial understanding, market understanding, through to, um, sort of readiness for production, we, we sort of honed in around, uh, uh, some discipline around that so that we don't, develop down a dead end or, or certainly at the pace of which there's technology, technological change that you start off developing something that by the time it's, it's ready, the market's moved. Yeah. Um, so, so there's, there's some, um, there's some discipline around required, but across the organization from commercial through to, to the production and aligning around something like that.

Chris Reichhelm:

Yes. The, the approaches and trying out, you know, within Paragraf, for example, did you guys get it right first time in terms of your approach? Or did you have to kind of iterate a few times and was there a system for that?

Tony Pearce:

I think we, we developed a system for that as we realized we needed it, I think the, the, and I, and I think that comes down to the way that in Paragraf's case, how, how the business evolved in that it was born from a process capability to be able to, posit graphene directly on substrate, something that nobody else could and still can do. Um, and that meant that, you know, we were, um, we were applying that to areas where we believe graphene could have the biggest impact in electronics. But also with an angle that, you know, let's not bite off too much more than we can chew in the first step. Let's not go straight for trying to create a processor or a logic chip. You know, let's look at something which is a demonstrator technology at the beginning. And we honed in on the magnetic sensing area and what is under the skin, a hall cross, which in electronics is pretty much the simplest device architecture you can think of. You can hone in on, but

Chris Reichhelm:

how long did it take you to get to that

Tony Pearce:

in that we, we had a, the first product within a couple of years. Um, so when the, the beginnings of the product was there before I started, you know, there, there was already work towards, um, what was, uh, uh, the first iteration. Um, but to your point about how we iterated it, I think what happened is we were, um, materials engineers, scientists, we were, we were separate from the semiconductor background, but. When you look at the, the use cases for magnetic sensing, it's everything from high physics laboratories to, you know, and cryogenics through to some of the markets we're looking at now around automotive, um, battery manufacturing and so on. At the beginning, none of us in the organization were experts in those markets. And that meant that the first iterations of the product, you risk them being either what it is you can make. Which isn't necessarily what the market is looking for, or, or what we think the market wants from a position of, of um, relative naivety compared to the people that you're speaking to. So, so the, the steep learning curve in terms of, okay, we, we've got, we know how revolutionary the technology is. We know the The differences and what performance this technology can bring, how do that, how does that materialize in the real world? What are the pain points that the customer really has where this product can have an impact? And then you redesign in some cases or modify the product, be it the, the size, footprint, the performance, the design. You tune it to that particular market. Yeah. So, so it went through a few iterations Yeah. On that basis. And, and now that simple product has, has created, um, four or five different flavors because we've realized that what is good for quantum computing, um, measuring, uh, at milli-Kelvin in temperatures, for example, in, in, which is one of the unique performance areas for graphene is a different product to what may be being used in battery manufacturing, where you're using it as a current sensor for detecting very small leakage currents and, and in that use case, sort of preventing, trying to prevent battery fires by detecting very small leakage currents. Those two products are different. And therefore, we've only found that out through, um, bringing in people that were knowledgeable in those areas, product experts, and gaining that experience so that we honed in on that. So some companies, of course, have that already on board. They may be developing a product where they are already the expert in that particular product area, and that isn't an issue. Um, from our perspective, You know, we have a platform which we're applying to different places, and that means you're not necessarily from the outset the expert in that place. Um, and therefore that leads to product iteration.

Chris Reichhelm:

Exactly. Exactly.

Tony Pearce:

So it was in the magnetic sensor. Yeah, yeah, yeah, that, in the magnetic sense of space, I mean, that was, it was a combination of it being, um, from a technology perspective, it was the right complexity of product and the right demonstrator, if you like, validation technology, validation product to show that the technology's real, we can do it in volume.

Chris Reichhelm:

And was that something that was, if you don't mind me asking, just digging a little deeper here, was it, you know, was it the commercial team or individual coming back and saying, hey, why don't we try this? Or was it technology saying we can do this. Let's see if there's a market for it.

Tony Pearce:

It's a little bit of both, to be honest. And I'd love to be able to say there was a pain point. We went out and fixed it. And you do that sequential? There was a little bit of both because, um, and that's going back to the point we said earlier about, you know, everyone sort of, um, living close to each other because those, those directional decisions therefore become a, a, a bit of a melding part of those different aspects. There's no point us seeing this, um, uh, market opportunity that is such a technological stretch that, um, that we're, you know, we're stretching too far. So, it is a little bit of both, but There is, if you want to bias it one way, it has to be commercial first, because you don't want to, particularly with a disruptive tech, is, when we talk about innovation and being disruptive, always on the positive side, right, that that's always a good thing. And on the whole it is right without, without us innovating disruptive technological steps, we won't make progress. But there's a, there's two sides to that coin. If you're too disruptive, if you are, if you, if you are, if you're coming in with something which is either incredibly left field or, or, or very difficult to incorporate, yeah. Um, then you. You may have bigger barriers to overcome to get your product

Chris Reichhelm:

and

Tony Pearce:

get your commercial traction.

Chris Reichhelm:

Absolutely.

Tony Pearce:

Um, and that's, so you have to measure the, if you like, the level of disruption that you hit first. It's a journey towards introducing something new and you break it into those steps. Not shoot the difficult thing first.

Chris Reichhelm:

It sounds like that level of coordination Between let's say these three different areas the tech this, you know The engineering and scale up and the and the commercial are just so critical at the early stage Do you and I'm not you know, and that is unless you've got that coordination, right? And you've considered those different aspects It's really hard to lay a big bet without knowing if you're going to, A, be able to scale what you've got, or find a market for it, or B, that novel with a solution. Um, so you've got to have those things answered for. It seems like before you really start motoring.

Tony Pearce:

Yeah, you do. And, um, we had a, uh, an added challenge in not 2D materials as a whole, but particularly graphene with that sort of bit of an image problem. Um, before we came along in that, you know, that 20 years that I was talking about, um, a great number of companies, big companies came into this, uh, joined them and like, came along into. Yeah, into this market space and failed. Yeah. And not because the material itself didn't, uh, or doesn't have the promise and the capability to do it, but it was manufacturing issues and, and, uh, there were, uh, there was investments lost in companies that have just changed direction or wrote it off their roadmap. Sure. So the, the importance for us was to, uh, and the reason why the first product was so early. Um, is, is that validation, that, that, that credibility that we, um, we can, and, and customers and investors alike now are still asked, come back to that question, how do you, um, how are you prepared, are you for, uh, repeatability, scalability, manufacturing? Um, and, uh, and that's how we differentiate. So, um, that additional, um, Motivation that we had, uh, and growth early strategy too.

Chris Reichhelm:

You had that, you know, the dog barking behind you or all of those failed graphene companies. I think we were, I think we may have worked with a couple of them over that period. So I remember that. You're getting back to that validation point a little bit. How, how do you go about validating? What is the validation process look like for you guys? And here I'm talking really about the commercial piece. Yeah. Because they may say, well, look, I love the idea of enhanced performance for these magnetic sensors. That's great. But I need to have faith that you guys can do this at scale. And that's kind of classic chicken and egg. So, you know, can you talk a little bit about what validation looked like for you guys?

Tony Pearce:

It's not that they could say that they do say that. And the, uh, I suppose in the early days when, when the main markets were scientific sort of instrumentation, I guess you don't get that. I mean, if you're, if you're. These days for markets, I said, like quantum computing, for example, they're not, uh, that's not a huge volume market yet. So people in that side are still, um, it's all about the science and the, and the immediate performance and, you know, the, the, the pressure of volume delivery is not there. But, um, whether it, again, sticking with the magnetic sensing, Um, I talked about that, that battery application. Now there we're dealing with, um, a, a market, which is, which is firstly, it's huge, um, and, and developing and moving very quickly. So the automotive industry has changed a lot in the last, in the last five years with the rapid emergence of electric vehicles and, um, whether that's a Tesla effect or, uh, uh, the, the, the rapid pace that China is, is developing its automotive Um, uh, industry, but, uh, it means that when you're, uh, you're introducing that disruptive tech, certainly before five years ago, it would have been almost impossible. The big companies, the Volkswagens and the Fords, at this point, well trodden, established new product development introduction process that takes years. Um, there are no, there were no shortcuts. There's a, there's a whole

Chris Reichhelm:

You've got to get in on the planning cycles and so on.

Tony Pearce:

Right. Which for a startup is, is, is, uh, is to challenge. EVs certainly shorten that a bit. There's a much faster development cycle time now, but it still comes to the same question. You've come along as a, as a scale up, uh, business and, and they're really. enthused by the initial samples, then that question of how can you ramp and how quickly, and they quickly talk about crazy numbers that you have to double check that there's not a decimal place missing somewhere. Um, and when you look at it, and you know what, you've got to, you can't just flag it and say, yeah, of course we can do that. It's just, scare us with the order, as you say. It's, it, you have to back it up with something, something incredible. And That is a, um, sort of, a challenge in the early stages because hard tech, particularly building a fab as we call it, you know, a manufacturing plant or, uh, in a semiconductor environment is incredibly expensive. And Paragraf has been very successful in raising investment to do that in the early stage, but even the money that we've brought in goes so far. And we have to walk that balance between being able to install enough of a capacity to demonstrate a pilot scale in the And by that I'm talking about, you know, hundreds of thousands or millions of units a year. It's not some small change.

Chris Reichhelm:

Yeah,

Tony Pearce:

but, uh, but not, we can't go out and build, um, a 200 million dollar fab to demonstrate that we've got 5, 10 years worth of, of capacity scaling done up. So, so you have to

Chris Reichhelm:

And are they saying anything to you like Are they saying anything to you? Like, hey, look, if you build this at this scale, it will come, you know, the business will come, or are they kind of holding back even more and saying, well, look, if you built it, then we know you're serious. Doesn't mean we're going to buy it, but

Tony Pearce:

yeah, there's a little bit of both and different, different, um, partners, different customers have different partners. appetite or requirements for that. Um, the, it ultimately is down to confidence in, not, yes, in your ability to scale, but that there is a, the right product fit and that your, your product makes it through the quali, their qualification into, into the end, end device. And yeah, if you don't get those initial,

Chris Reichhelm:

how long does that, and how long does that qualification process take?

Tony Pearce:

I mean, to get to the, in automotive, there's still, you know, you're still looking at between sort of four or five years to get to the, to the high volume stuff, but there's significant volume on the way. And that's the, that's the difference is that, that even in the, uh, one year in, two year in, three year in during the qualification phases, there are still challenging volumes in those stages too. So it's not a digital step where suddenly it goes from zero to tens of millions. There are, you know, there is a little bit of an opportunity for you to grow with it, albeit pretty steep. Um, and with, with semiconductor devices, when you're at the device level, when, and thankfully, you know, die size is a relatively small number of devices you can get from a wafer is quite large. So the gearing is is in your favor. Um, if you make the right supply chain decisions and use outsource partners for some of the non core stuff you can ramp up quickly um and it's as much about credibility in that supply chain management and credibility in in your your ability as a business to To manage an organization that's got that kind of production demand, um, and that you're not just, um, the rabbit in the headlight scientist that, uh, is, uh, and, and is gonna, uh, doesn't know and never been in that position before.

Chris Reichhelm:

Absolutely. What are those, you know, when you're, for argument's sake, you've got automotive, OEMs very interested or tier one, tier two suppliers, very, very interested in what you've got. And they want to start talking about integrating your solutions into, into their products. Um, and, uh, and so you're having conversations with supply chain partners. What are the things that they are looking for? You know, what do they want to, what's going to give them confidence? That you're not just a group of zany scientists who've never even considered something like this. What needs to be in place?

Tony Pearce:

You need to, I mean, there's the stuff I talked about earlier about our ability to be a credible manufacturer, but And there's a, there's a, there's often a gap, um, between the, the disruptive product that you've manufactured, say in our case, this magnetic sensor, and it's the same for the, um, the molecular sensor, which was the follow on product that we now have out there. Um, that you need to help the customer to know how to implement that product into their systems. And there's what, um, my colleague Tom calls the implementation gap, which is that when that, uh, that customer, whether it's a, uh, a big corporate or not, they look, they, they like, they understand the, the product they have. tried a few in their R& D environment, but then they have the challenge as to how, how does that product get incorporated into our product? What's the, what's the gap, you know, to how we implement that? Because there's rarely such a thing as a drop in replacement product unless it's, unless it's coming in at just at lower cost or it has a better environmental footprint or whatever it might be, and there are some cases where it's just a drop in. But for most Sort of innovative technology. There are, um, there are associated developments that need to happen to arrange your core product to get yourself adopted. So, uh, and what we, uh, what we've had to do is, is in some cases, bridge that implementation gap to perhaps develop a little bit further around our own product that's not core. You've got to be careful that you don't go too far down that line and, um, and, uh, and get swallowed up by it. But, but to, uh, for example, build an associated sort of, or ASIC to go with, uh, with a sensor or to build a module, um, that the sensor goes into or to build some apparatus that allows you to demonstrate the performance of the. Sensor that the customer doesn't have to work it out for themselves. Yeah, because then it can be well you, that you can get a little bit of that, that sounds quite difficult. I'm interested and I'll come back to it when I've got time. Yeah. So you've gotta, you've gotta help them to make it, you've gotta make it as easy for them to, and make it as easy as possible. And sometimes that means doing more than you originally intended to do in that market area at a demonstrator level, um, which then helps them bridge that gap and then they can run with it. Once they start developing around your product, you're in, um, that's when you know that, you know, they've bitten and the momentum really starts.

Chris Reichhelm:

How much of that is extra kind of product development work or solution development work? Uh, and how much of it is. Really great communication, relationship development, management, I don't know, education. Certainly,

Tony Pearce:

yeah, we certainly had to have a lot more application engineering, if you like, um, and Things around, um, software and, and, um, in some cases, sort of additional instrumentation and things like that, which we know, you know, we have to bring people in there and work with partners to develop that, um, exactly for that reason, for, for, for bridging that gap. Um, you, you certainly need the, the close understanding of, of the customers. Um, internal sort of dynamics as to how they bring a product to market and what, what to understand what they need, what the difficulty is to make it as easy as possible. As possible for them. Yeah, not all companies are the same but but often you're you they have the same Internal challenges that you have they're really excited by the technology. They want to find a route to get it into their product They've got budgets. They've got bosses that they need to convince, right? So You you you build that face to face Face to face relationship with that customer, understand what they need and help them on that internal process within their business, because ultimately you've got a shared interest and a shared goal to get that technology into the market.

Chris Reichhelm:

Are they looking at your scale up or manufacturing process as well? You know, is it, is there a, uh, I've heard this from others, uh, from clients and the like, but there's. You know, there can be issue with novelty in manufacturing processes, um, because they're not immediately accessible to others, right? They're novel. Yeah, true. Whereas if you're going for something that's at least recognized. and ideally modular as well. It helps with, you know, integration if you need to do that later on, or it's just recognizable by other partners. Do you find that?

Tony Pearce:

Yeah, yeah, absolutely. I mean, that, that, that consideration is also one of the core, um, I suppose, principles or strategies that we've had within, uh, within Paragraf to keep the manufacturing process as, as recognizable and as industry standard as possible. So while there is. You know, closely guarded IP around the process of manufacturing graphene and it doesn't end at the graphene. Graphene is a single atomic layer of carbon and trying to work with the materials around it to even put some metallic contact onto a single atomic layer is no mean feat. So there's a whole lot of IP and know how around the process. But if you were to walk into. Um, uh, our facility, which, um, you should come down to Huntington, Chris, and see what we've, uh, what we're doing with the new build. Um, you'll, it'll, it, you'll recognize it as a, um, as a, a semiconductor fab with the same kind of equipment in that you would see in a conventional silicon or compound semiconductor facility. And we've deliberately made Um, made that approach because, you know, I've been, I've tried to, um, introduce novel processes to mainstream silicon industry for a long time, and, um, the, the industry is, well, while it's incredibly innovative, it's also very reserved about how it works. How it changes and what it adopts as a new material or a new process step. So if you can de risk that by having recognizable tools, OEMs that they know, so that they don't have to completely rip out incredibly expensive tooling and bring in something different. So we're able to give that confidence level to a degree, but we have to balance that between. Certainly keeping our IP very close. Um, yeah. And, you know, in the future then there's, you know, there's, there's clearly licensed models and things that we can use. Further down the line, but you can only, again, you can only do that if you're industry compatible. If you're not this crazy, crazy science experiment. Exactly.

Chris Reichhelm:

Exactly. Um, you and I have talked about this next question a bit, um, how, what's it like trying to build this in the UK at the moment in 2024? You've done a lot of building of semiconductor facilities and operations in the UK in general for the last 25 years. Um, uh, longer, I think. Um, so what's it like doing that in 2024 with a company like Paragraf?

Tony Pearce:

Yeah. Uh, it is challenging everywhere, right?

Chris Reichhelm:

Yeah. It is

Tony Pearce:

particularly challenging in, in, um, at this, um, uh, at this time in, in the uk in terms of being able to, I said about the, how capital intensive, how expensive it is to, to set up a, uh, a facility, hours to set

Chris Reichhelm:

of interest. How expensive is, how expensive is it? to fund a plant.

Tony Pearce:

What's

Chris Reichhelm:

up for you guys? You know, what's

Tony Pearce:

the rate? We're not doing everything in one go. What we're doing is building a modular approach such that we can, we can build out at the rate that we can both afford and that the demand dictates. Our facility in Huntington is you know, 40,000 square foot Um, building, um, and that's in the tens of millions, um, of capital expenditure, um, over 50 million, um, to do that, depending on, um, uh, the exact, uh, specifications and, uh, and details of what goes in, but, but it's, it's incredibly expensive. And the, now that's the same anywhere. That's not a UK problem. It is. Yeah, exactly. Exactly. Exactly. And all around the world we have that. You can see that the, uh, the equipment cost is, is what it is. What we, what we found that the, both the, some of the cost difficulties and the rate at which we can expand is not necessarily All right, just around the expensive tooling that goes into the fab, um, but the fab itself, um, the, if you are a spinout from a university and your, your software or, or, or some of which is predominantly desk based or small lab scale, you can go to an incubator into a science park and there, the UK's got quite a few of that, Cambridge especially has, has a great science park infrastructure, which is built around that kind of, you know, Um, sort of early stage, but when you, we had to leapfrog that from the very beginning because the, the, the amount of footprint required the type of facility that we needed, meaning that you have to go out and find, in our case, a big empty shed, which was a, which was a, an old door factory in the outskirts of Cambridgeshire, which had, which was gutted, um, and turn what was, I mean, um, Famously, it was raining and snowing more on the inside of the building than it was on the outside when we first took it over. But create from the ground up something. There's no The only thing we speculatively build in the UK is distribution warehouses, right? There's no, there's no, uh, ready made factory or You have to take on something that's either old and obsolete, um, or is not designed to be a factory, is designed to Yeah. Be it FedEx or Amazon, which is, um, so, so that, so we've had to, those sorts of facilities don't have the power that a semiconductor facility needs, doesn't have the infrastructure around it. So we've had to, to go through, um, don't start me on planning, but, but we've had to go through sort of various, what I, I feel are sort of UK related issues around getting the bureaucracy and the cost and the time around getting. a local sort of electrical substation connected and power in the ground. And when you're venture capital backed, your investor doesn't want that money going into wires in the ground. They don't want to be building UK infrastructure. They want to see technology. They want to invest in the product. So when you your business model includes a significant amount of Um, electricity, power network works and water treatment and things like that, for example. That's not a sexy investment for a lot of people.

Chris Reichhelm:

But it's critical, because without good infrastructure, you can't do much.

Tony Pearce:

You see that? Um, you know, when we look outside the uk, there are, um, there are, there's certainly, um, uh, support, um, in, in the US or in uh, uh, recently I saw an example in Germany where there is government support for those kinds of infrastructure, um, challenges that scale up companies have. Where we focus on what we're good at, which is, uh, which is the, the, the scaling the technology and getting the products to market. Um, we're, we've been good at attracting investment towards that, but we get some support to give us the, the footings to enable us to build upon. Um, we're not expecting there to be a ready made semiconductor fab sitting there empty for us to move into. I mean, that's, that doesn't happen.

Chris Reichhelm:

Yeah.

Tony Pearce:

But if we, if you're able to, yeah, exactly. Or if we identify somewhere, um, that, that looks great, then, then some Um, some support to help us build that basic infrastructure to start from would be, would be really helpful for accelerating, um, accelerating our progress.

Chris Reichhelm:

Yeah.

Tony Pearce:

Um

Chris Reichhelm:

What about talent?

Tony Pearce:

This, I knew you were going to ask that, Chris. The, the, the, this, again, that's not just a UK problem, although we have got some sort of key shortages in key areas.

Chris Reichhelm:

It's not, yeah.

Tony Pearce:

Um, the In semiconductor electronics generally, um, take a look at what's happened in the U. S. with the, uh, the CHIPS Act. They're building fabs certainly faster than they can staff them, um, and the skills talent shortage, uh, is, is recognized, um, uh, across Europe and the U. S. as, as, as difficult. Because we've, and certainly in the U. K., there's like a, I look at it a bit like a lost generation in semiconductor manufacturing. But, uh. In the eighties, nineties when we had lots of facilities around Scotland and, uh, Wales. Wales and, and, and in particular, um, there was quite a, uh, there were clusters and, and an ecosystem skills and people were relatively transferrable and, and you could put a, um, uh, a wreck. We still got, um, um, general sort of science and engineering people around, but there was no, now nobody would drop in set sort of semiconductor or electronics experience. So, so, we are recognizing that. I think in the UK there is the beginnings now of courses to bring people into the industry again, and it's starting to gain a little bit of momentum, but there is a

Chris Reichhelm:

And specifically on the manufacturing, on the fab side. Yeah,

Tony Pearce:

yeah.

Chris Reichhelm:

Not just on the design side, which we've had for a long time.

Tony Pearce:

But that, I think that's covered. I'm a bit biased because I'm, you know, but I think, you know, there's a, you put semiconductor out at the time and you get chip designers and, and, you know, that's, that, that doesn't help with, with, with our end of the, of the, of the the industry. So those sort of, um, those practical, hands on manufacturing skills are great.

Chris Reichhelm:

And it's also, it feels like an integral piece, it feels like an integral piece for innovation in general. Because it's what, you know, design is super important. But, if you understand the process required to make, through that.

Tony Pearce:

We'd have to do our part. Yeah. We, exactly, we do our part in, in terms of trying to, um, bringing that. Uh, new and, and certainly younger, um, scientists and engineers through, we're doing for our size, we're doing sort of a disproportionate amount of internships and yeah. Um, so, uh, and, and, uh, starting now with more apprenticeships and things like that, which is, uh, we found incredibly value. We pulled in interns from the US and MIT and and and so on who have come over to Cambridge and. And work with us for a, either a summer or in some cases for a 12 month placement. Um, but it, it, that's an investment that's going to take time to pull through.

Chris Reichhelm:

Absolutely.

Tony Pearce:

Um, and, uh, but, and certainly there's a, there's a much, uh, better, more diverse workforce coming through at the younger ages now. There's, uh, again, that's taking. Taking time to make it up into the more senior, uh, experienced positions, but we're, we're, we're getting there.

Chris Reichhelm:

Yeah. If, um, you know, to manage that to you, and this isn't just a question for you guys necessarily, but should companies be considering broadening their manufacturing base beyond just the UK, you know, to the US, to Europe, to Asia, in order to, in order to meet the demands now?

Tony Pearce:

I think the international scaling is. is important. I think we're in Paragraf's position is that, you know, we want to to stay in the, in the UK and build a company in the UK, um, while those sort of attractive conditions exist. It's, without doubt, there are, um, in other regions, incredibly enticing conditions, be it, um, uh, initial incentives, but, but, you know, you have to operate in a country beyond those initial incentives that, that, you know, the help to set up is, is really valuable, but you're, at the end, you're still left with a factory. You have to operate in that. There are other considerations too, um, but, uh, I, I think where, um, where we see the future going is, is certainly remaining UK biased, but we also need to look at where our customer base is, where the, uh, where the right partnerships are. And as we. As we move into higher volume markets in automotive or other related industries, then it may well be that we have other locations in the U. S or Europe that complement that. So we set up Paragraf USA quite early on, made an acquisition in San Diego, which is acting as our, our first sort of. Modest footprint in the US, um, but, you know, already through that, you know, we see, uh, you know, there's an attractive business environment out there, which is, um, which is, which is working at the moment, complementary to what we're doing in the UK.

Chris Reichhelm:

Yeah. One final question, Tony, if I may. The, um, you've had a great career, a long career in the industry.

Tony Pearce:

It's not, it's not over yet, Chris.

Chris Reichhelm:

It's definitely not over. But if you were to, if you were to stop and think about some of the really big lessons you've learned. Uh, that you'd like to see applied more or wrapped into the thinking, particularly for these types of businesses as they're setting up. What might they be??

Tony Pearce:

I think I'd probably touch on the points we talked about earlier, I think from, because even um, in the, in the last few years in the Paragraf Environment, I've learned lessons in what's important in this scale up. Sort of lab to production journey and we talked earlier about that getting that market product knowledge and credibility with those customers that that was certainly, um, a very quick lesson within that environment for us at the beginning is to make sure that. You are able to, to be an authority in that market and talk, uh, directly to the experts and not come in as, uh, as a interesting science, um, That, uh, doesn't really understand the market, uh, doesn't really understand the product. You've got to be able to answer questions directly and understand their pain points and be credible in front of the customer. So for, for any startup scale up business that, um, uh, if you have to bring it in and do so, bring in that market expertise. Um, early, it will, it will short circuit, it will short circuit a lot of, um, revision of your product and development time with your product to, to meet the customer's requirements. Um, I'd say the other lesson was that, again, repeating what I said before, that implementation gap idea. Be prepared to, to have to develop around your core product to help your customer implement your technology. Don't expect them. To do all the legwork for you, help them to get your product to market because, uh, you know, you're on that journey with them and you share a mutual benefit in being successful. Don't just expect them to work it out on their own.

Chris Reichhelm:

You've got to be budgeting for that. That's part of the whole thing.

Tony Pearce:

Yeah, yeah, yeah, yeah. Cost, size, resource wise is different expertise. Um, so you, you do have to think about that. Um, I think finally, probably from a practical perspective, when you're, when you're building, physically building the company and the infrastructure, and that's to make sure you design in flexibility in your capability going forward. Building a factory, building a fab, ordering equipment, buying in, Long lead items. There's a timescale involved in doing that. It's longer. It's a slower heartbeat than the way the technology is moving. Yeah. So, you don't design a facility or a factory or a capability around just around what you think you need today. Build in some flexibility in that. Don't underestimate the ability of the scientists to completely turn a process on its head within three months and to find a whole new way of doing something, which is much better. And, and that stuff I put in place that I ordered six months ago, we don't need it anymore. Um, so, you know, you have to, um, you have to be flexible with that and, uh, and be prepared for some, Um, some, some sort of changes of direction, not just when that happens in your strategy or your product development rolls down into your facility and your factory and your, uh, and the practical infrastructure that you need. So being prepared for that, um, potential change in the detail, um, when there's big money involved, um, you don't want to be building, uh, building something which is, you know, Obsolete before it's finished. So, so, you know, consider, um, try and be predict the unknown when you're building the factory as well.

Chris Reichhelm:

I have to ask, I can't let this hang. How do you build in that flexibility? Like, do you create contracts where you can give it back or undo what you've done, if you've already started building?

Tony Pearce:

Yeah. Equipment manufacturers are not great at the return model. Um, no, but, but luckily, um, I think that you, you don't wanna become it. It fits with the, with the strategy I said earlier about not being too sort of specialist or, or, yeah. Wacky in your manufacturing process.

Chris Reichhelm:

I see what you're saying. Yeah. I get, I get what is, is

Tony Pearce:

make, make the tools as relatively generic as you can, such that, yeah. Um, if you can make process changes, use different materials, chemicals, um, such that the changes, um. Are in the recipe, not too much in the throw the tool out and buy a new one.

Chris Reichhelm:

I get it. I get it. I get it. Yeah. That makes sense.

Tony Pearce:

That's it. But sometimes when you're trying to keep to a budget, you strip out all that stuff and you go, I only needed to do that, so I'm only going to buy it to do that. And then when the process changes, it doesn't do anything else. So sometimes it involves having a little eye to the future.

Chris Reichhelm:

I get it. I get it. Tony, amazing. Thank you so much. This has been, this has been great. A nice deep dive into thinking how you have a chance of taking one of these things from kind of very early doors through to something that's far more mature. This has been wonderful. So thank you so much for joining us.

Tony Pearce:

Thank you, Chris. My pleasure.

Chris Reichhelm:

You've been listening to the Lab to Market Leadership podcast, brought to you by Deep Tech Leaders. This podcast has been produced by Beauxhaus. You can find out more about us on LinkedIn, Spotify, Apple, or wherever you get your podcasts.

People on this episode

Head Shot

Chris Reichhelm

Host