#010: Andre Brown - Principal Engineer on Amazon's Transportation Sustainability Team - Digital vs. Physical Prototyping, Creating Flat Hierarchies within Teams, Innovating Faster, Designing in Virtual Reality, Eggs and Vacuum Cleaners, Procedural Design.

Digital vs. Physical: When to Prototype—and When to Simulate
Andre Brown, Principal Engineer on Amazon’s Transportation Sustainability team (and former Shark/Dyson R&D lead), joins Rob to unpack the art of knowing which problems need cardboard, which need CFD, and which belong in VR. From drag-cutting truck fairings and crowd-sourced race-car CAD to Dyson’s “egg-inspired” robot vacuum, Andre shows why flat hierarchies, cross-discipline sketch-offs, and procedural design tools accelerate innovation. If you juggle human factors, aerodynamics, and sustainability, this episode is your playbook.

Transcript

SPEAKER_00: 0:00

you know you can't replace prototyping so like the different roles that i've had in the past that prototyping is is fundamental to some of those roles and then simulation is like the death of prototyping in other roles because the simulation is coming on so fast and It's just, it's been interesting for me having a breadth of roles to be able to say like, okay, in this case, like I actually do need to prototype and I need to hold the thing and I need to look at it. We need to paint it and we need to give it to a customer to use it.

SPEAKER_02: 0:45

So, hey guys, what is up? Rob Irwin back again with another episode of Designing in the Wild. I have been playing schedule tag with Andre Brown for a little bit and super excited to finally get a hold of you, Andre. How have you been?

SPEAKER_00: 1:02

I've fully recovered from a broken hip, but other than that, I'm all well, yes.

SPEAKER_02: 1:09

Oh, my gosh. From all the activity we were talking about earlier you were doing with the new store and the kiddos, it sounds like you're putting the new hip to good use.

SPEAKER_00: 1:19

I'm getting old. As soon as you start breaking hips, you know that you're getting old.

SPEAKER_02: 1:26

Exactly. Well, for everybody out there who doesn't know, Andre and I worked together at Amazon back when we were designing the Dash card But he is now the principal engineer for Amazon and Transportation Sustainability. He's the chief engineer for innovation over there. And you've risen the ranks since I've known you, Andre, for the past three or four years when we were working on the Dash cart. What are you guys working on right now, if you could talk about it at all?

SPEAKER_00: 1:57

So I am working on trying to make our middle mile transportation fleet more efficient and sustainable with a specific focus on trying to enable zero emission vehicles. And most of my focus initially has been on the design of the aerodynamic components that go onto our existing fleet. It's kind of a short term, try and improve your current product, by making it more efficient, use less fuel. You know, as they move to alternative fuels, efficiency, you know, fuels cost a little bit more. The technology costs a little bit more and they need to get those sunk costs back again through efficiency improvements. So, you know, it's about reducing the amount of fuel used, which reduces the cost, which reduces the carbon emissions. And so, you know, a lot of simulation and working with suppliers to get aerodynamic parts fitted to hundreds of trucks that are going out every year these are new trucks and retrofitted trucks

SPEAKER_02: 3:08

so right yeah i'm guessing there's a challenge there with the different models and makes and whatnot for compatibility

SPEAKER_00: 3:14

there's we we typically um you know buy from more than one source um it's it's a common practice and it just means that you can scale really quickly And so certainly with Amazon, fleet is growing really, really quickly. And being able to have two different trucks or three different trucks just allows you to go at the speed that you need to go up. But it does present problems when designing components for those trucks because every cab has a different geometry, every chassis is different, all the boxes are different, you've got different... um so you've got different truck oems different box oems different trailer oems and then add on top of that you've got um component suppliers which are typically smaller suppliers that are trying to design uh you know skirts and fairings and devices that reduce drag and then so you're layering all of these different um variations and and so you have to have a really very, very lucky with AWS that we're able to run all the simulations in-house. So we were doing aerodynamic CFD simulation and we can take all of these different various designs and then do what happens if you run with fairing A versus fairing B and what happens if you run on this route versus that route and you can do a big design of experiments that gets you to some really meaningful data that you can get really quickly because we've got this amazing compute that scales to whatever you throw it at. We're running with hundreds of millions of cells.

SPEAKER_02: 5:03

That drives down your analysis time as well. I'm guessing given the economies of scale for distribution and different types of trucks, you're able to retrofit like as you mentioned different iterations of different fairings and different designs across all the permutations and come out uh on top with the with the best

SPEAKER_00: 5:27

we we're we're doing simulations now with with outputs of 200 million cells and we're running it on you know over a thousand cores um in parallel so that you know these are massive supercomputers and and we can run you know i'm running half a dozen different geometries all in parallel. So, you know, you've got, you've got this, this massive resource that you can switch on and we're doing simulations that would have taken a week, you know, to run just one simulation. And we're doing, we're doing half a dozen overnight in 12 hours. So, you know, this is just, just the answers that you can get and the fidelity that you can get in the simulations. It allows you to, um i would say it allows you to to skip the step of uh prototyping to try and work out what you want to do um and you know this is a general theme of mine i've been involved a lot in prototyping and simulation uh you know and In some cases, you are able to skip that initial prototype step and go straight to something that you can go and test on the road, real data. And in other cases, there's absolutely no... You can't replace prototyping. There are different roles that I've had in the past. Prototyping is fundamental to some of those roles. And then simulation is like... the death of prototyping in other roles because the simulation is coming on so fast. Right. It's just, it's been interesting for me having a breadth of roles to be able to say like, okay, in this case, like I actually do need to prototype and I need to hold the thing and I need to look at it and we need to paint it and we need to give it to a customer to use it. And, you know, my current role, it's more the opposite. It's okay, let's do a really high fidelity simulation and skip the step of prototyping and go straight to okay, let's deploy to 600 tracks and collect real data for how they perform in the field and then get better again for next year because you know there's another 600 coming next year or whatever the number is.

SPEAKER_02: 7:52

Yeah, that's super interesting. I know that I definitely agree with you. There's definitely some hands-on involvement with prototyping. I think a lot of things that require or concepts, features that require hands-on prototyping may be skewed towards maybe the more human interaction and behavioral stuff, maybe ergonomics and human factors. And you're talking about fluid dynamics in air and just geometries on automobiles And so that definitely makes sense to to spit those out quick and get your answers quick. Now, I'm curious, are you using off the shelf like fairings and things like that, or then or are you then going to model in customs and have those built and tested?

SPEAKER_00: 8:38

So it's a bit of a new muscle for Amazon, really like that. The transportation division is relatively young as a as a group within Amazon. And the pace at which it's growing has meant that they've relied on OEMs. So from a design perspective, initially, we're just leveraging existing supply chains. And in the very first deployments, we're saying, OK, what can we get that catches the procurement schedule? And so we're analyzing what's on the market, and then we're making some selections, and we're learning from that. At the same time, we're looking to influence the design for the following year. So working with those same suppliers, we are looking to say, okay, well, we've got this great simulation capability and we've got some ideas for different shape fairings, or we want to use some machine learning to help us iterate the form to come up with a lower drag shape that might not have been conceived by a designer. And then working with those suppliers to say, okay, let's prototype and test in a small batch trial for the following year. So it's a little bit of both, but we're not at the moment trying to run before we're walking. And we're not standing up a whole facility at this stage for doing our own thing. more about like, let's go quickly and influence what we can buy today and try and improve what we can buy today. And then, you know, see what we learn and how we can apply that to the future.

SPEAKER_02: 10:27

Yeah, I was always enamored with the speed at which Amazon moves. And, you know, I know that there's a lot of corporations and companies out there that try to move quickly. And this is this brings me to a thought about ambiguity in the design process. And I know that, you know, with with these simulations it really helps to funnel the the questions uh quite quickly and and i know that you have some history with uh previously working at shark and let's make some work at dyson and such and and you alluded to the prototyping side of things on the flip to kind of running simulations i'm wondering if you could talk about how you in your mind balance uh the level of ambiguity across a project well let me rephrase it like the level of ambiguity with uh let's say a vacuum design vacuum cleaner design i know you can engineer uh simulate the flow analysis uh and all sorts of a number of other things but when it comes to like the behavioral rasp behavioral aspect of of developing prototypes in maybe a realm such as Shark or new R&D there. Talk to me about kind of how you think through balancing that and why is it important to kind of get these models in real world aspects quickly?

SPEAKER_00: 11:51

So, like, in retrospect, having worked in, you know, through no... no specific choice i just kind of ended up working in the vacuum cleaner industry like i had studied automotive engineering and went for a job at dyson and they they were talking about doing an electric car and i ended up doing robot vacuum cleaners and that was really my introduction to product design right i didn't study it at university or anything and i i i'm absolutely amazed looking back on it at how um multi-disciplinary that you know just designing something as simple as a vacuum cleaner is and and it it really is like the complexity of it when you're looking at trying to um to to innovate or to improve the the state of the art um you've got so many factors that come and compete for your attention as a as a know when you're in the ideas phase so some of those things are like hey i want to make sure that i'm you know clear of ip that is pre-existing or that i create ip right there's there's a whole um you know what can you do what what can you create that's protectable um there's That's one set of pressures that impinges on the design process. The second is you've got this product team that is trying to define a marketing message and has got a whole set of criteria about what will and won't sell. They've got extensive experience. They've got supplier relationships, big box stores that are going to say, okay, well, you've improved it by 18%, but really, that's nothing, right? Like, we can't write 18% better on the box, you have to get to 50% better, right? So like,

SPEAKER_02: 13:57

yeah, market shifts. Yeah,

SPEAKER_00: 13:59

yeah. So you need, you've got, you need to think about your new designs based on like, trying to meaningfully improve some metric or another. And then, and then you come back to the question of like, which metrics are important, right? And so, There's some metrics that are important from an engineer's perspective that then knock on, define the shape and the air path and the motor and the things like this. So you might be like, okay, we need, you know, marketing wants to be able to write the most powerful, xyz and so you need to be able to deliver that from an engineering perspective and that

SPEAKER_02: 14:39

right

SPEAKER_00: 14:39

that means that your air pathways need to be low pressure drop and that means that your motors have to be high efficiency and they you know they want to have a long run time so they need to use less power per you know we used to call it air watts but it's it's kind of a measure of the air power and But you can't have too many batteries because it gets too heavy. There's all these competing technical things. And then layering on top of that, there's all of the user of the product, the person that you're designing it for. They care about some aspects of the design and they really don't care about other aspects. And so you get into the questions of like, While this is a cleaning product, it should probably make sure that it does its primary purpose really well. But it needs to be usable. It needs to be ergonomic. It needs to be lightweight. It needs to be beautiful. Especially vacuum cleaners are starting to be left out and about more so that there's a whole shift towards convenience. know these softer kind of metrics like how do you how do you define convenience well you probably have to go and make some things and you know talk to some people and let them try it and that forces you to prioritize some of those um you know harder to define um metrics and i think for me that depending on which of these different areas of focus you're you're looking at any given time you might make a different prototype um or a different simulation in order to get after improvements in each of those areas and we used to make dyson was a fantastic example they they have this um approach to prototyping that forces it forces their whole industrial design language really. And it all is driven from an unwritten approach to prototyping where they ask you to make it in cardboard first, right? No CAD. Right. If you've got an idea for anything, go and make it in cardboard and it's fast, it's cheap. It's really accessible to any designer. It doesn't require lots of tools and equipment, right? It's like a scalpel and some glue and, That's it. They developed techniques for making radiused edges. The intricacy that I've seen in the Dyson prototyping labs for cardboard prototypes is absolutely incredible, to the point where they're making cones, and they're making spheres, and they're making cylinders, and chamfered edges, and all sorts of detail. The interesting thing with that is that it forces the design language to be based on primitives. And so you'll notice that Dyson has a very distinctive design language, but it's not a top down, like, here's a design Bible, you shall do this and that. It's driven by the approach to prototyping, which I find absolutely fascinating. I mean, Dyson is a very closed company. They do a lot of the development internally, a lot of user testing internally. Shark was the absolute opposite. So Shark was like, okay, we're going to go and ask our customers what they think of any prototype, right? So at every level of prototyping, right from the earliest, we've got some sketches, right, like pencil sketches. Maybe they've been, you know, cleaned up a little bit and, you know, they've had some line work done on them and they've had some, you know, not computer renders, but like artistic renders. We go and talk to people with a set of images and a set of things that we might write on the box about performance or ease of use or any particular feature. And you go and test those with people and ask them their opinion. And that within Sharp would drive a lot of the should we do this or should we do that was driven by the voice of the customer reacting to early prototypes at any You know, everything from sketches through phone models, through, you know, painted ID models, through, you know, we start getting into working prototypes. Those working prototypes would also be tested with consumers. They would be, as they get closer to production, they would be left with consumers in the home for 30 days, you know, like just getting a proper read on what they saw as valuable products. You know, features would be ranked and you'd almost ask the customer to say, well, what would you spend your money on? Would you like feature A, feature B, feature C? And you get a really good handle on how the product's going to perform in the market off the shelf, how it's going to perform in the home, what customers are going to like about it. And that's all enabled by prototyping. simulation can't do any of that right like it can't get you to those kind of softer um insights but simulation allows you to confidently predict the performance before you start prototyping so i see it as a way of digitally prototyping to gain confidence so that as an engineering group you can say to your product team Well, yes, we think we can deliver 50% more suction or 20% better cleaning. And you've got to be able to say that with confidence, because when you're in that really ambiguous early stage, you can't just promise the world, right? Marketing will want everything, right? Do you want more suction? Yes. Do you want lighter weight? Yes. Do you want better cleaning? Yes. Okay. So at some point you have to say, we think we can deliver this for this price. And that's where simulation comes in early. And we used to use simulation a lot for both looking at the layout of components and the air pathways between them, but then also fine tuning things like We used to put batteries in the handles and things like that, right? To try and put the mass where, you know, really close to the hand to make it feel light. And when you do that, then you have temperature issues to deal with. And so we used to use simulation to try and make sure that there were no hotspots on the handles and that it was at a comfortable temperature. And that would get, you know, you would build test rigs to go and test with customers to confirm the simulation and stuff like that. I definitely, the vacuum cleaner industry and any kind of technically complex consumer product where the customer has to interact and use the product both as an ergonomic, but also has to have some kind of performance. Like you've got to do all types of prototyping and simulation all together at different phases in order to get the answers that you need at the right time.

SPEAKER_02: 22:12

Yeah, I think that convergence of the ergonomics and the human factors with the internal engineering and innovative, like integrating that form factor that is functional at the same time as being obviously usable to the customers is a challenge for a lot of teams, I think. In my opinion, I've seen and worked with teams where you have disparate competing evidence to converge on a product. And oftentimes, depending on how it's weighted, right, I think that there's different weighting metrics to each one of these as well, right? Yeah. And I'm wondering if you can speak to how, you know, in a lot of organizations, you know, engineering is over here, design is over here. They kind of create in their silos, if you will, and then kind of have touch points and milestones where we all make a prototype together or something and see what's working and what's not. But oftentimes you come up against, I'll call it a clash for lack of a better word, between a competing performance metrics, whether it's the handles ergonomics, but the weight needing to be a certain weight or distribution at the same

SPEAKER_00: 23:35

time.

UNKNOWN: 23:36

Yeah.

SPEAKER_00: 23:36

I've been so lucky in my career that I have not worked in those silos. And the companies that I've worked at when I've been doing product design have been acutely aware that even to the point of job descriptions, your job description was design engineer, right?

SPEAKER_02: 24:01

If

SPEAKER_00: 24:01

you came from an art background, like pure art, fine art, illustration, if you came from a pure engineering background, it doesn't matter. Everybody was design engineers. And what that did was firstly, it flattened the hierarchy so that there was no us and them. Everybody had a common title. And secondly, it empowered everybody to say, well, actually, my job is to influence the engineering, even if you've come from a fine art background and vice versa, right? As engineers, like I've, from a personal perspective, I've struggled a little bit with, I think, making beautiful things. And I have my own sort of, like, I have an attempt at making something that,

SPEAKER_01: 24:55

you

SPEAKER_00: 24:56

know, like when I look at it in the computer or like I make a prototype, I have my own, aesthetic judgment, but like when I see somebody who's actually got some talent, like do the same job or, or, or takes, takes my underlay. And, and then you see people make things that are so much more beautiful than, than, you know, it's just a different skillset. Right. So it, it means that like every job is every, everybody's responsible for all aspects. Right. And they, and it empowered everybody to speak up about like, this should be better or that should be better and you never had like that's not my job or like you know my job yes so i i think that's been immensely um powerful um and yeah and and also like the co-location of people like i've been lucky in that that as engineer like i come from An engineering perspective, I might be looking at the arrangement of the internal components, right? Like for best weight distribution or like for best airflow through the device or, you know, things like that. Working like where you've sat next to the guy that's designing the skin, for want of a better word, like the shape and the part, you know, the color breaks and the part splits and, you know, even getting down to like, where where are we what are we doing with the screws and you know all of that kind of um the design part of it like being able to sit next to the guy so that you're sharing the same cad even at the underlay stage and you know you can come and sit in on all of the sketch hacks where you you know you're taking the underlay and there's 50 000 sheets of paper of the underlay printed and everybody's sketching on top like it it's been really really nice to be part of an inclusive product design. And product design isn't just about one discipline. To have a successful product design, you've got to have engineering and legal and marketing and industrial design and all of this. All of the teams that I've worked on have been very, very small design teams of less than six people. responsible for a product, even if that product then is making hundreds of millions of dollars for company XYZ and being manufactured at huge scale. And when I say it's half a dozen people, I'm talking about the early product definition phase. Obviously, there's a much bigger team that is needed to get things into mass production at a later phase. And there's a lot of the design a lot of the design work happens in those later stages. So this early team that is responsible for the product vision and getting it onto the roadmap, those people stay with it to ensure that the vision is seen through into production. But the touch points with the design become more about course correcting as you get closer to mass production rather than wholesale changes in the layout or the form or the design, which happens early on. But yeah, I've not experienced the silos.

SPEAKER_02: 28:35

There's a lot to unpack and a number of points you made there. I want to back up just for a minute about the comment you made about the flat teams. This was definitely prevalent And at Amazon, when we were working on the Dash card and we went from, I think when I was hired, it was myself and 12 engineers. And then we ramped up in a year to 70 engineers and multiple facets of teams broke out with firmware and software and hardware. But we all worked on the same floor in the same giant open room. And you could walk over to someone's desk and say, hey, Andre, what are you working on? Check this out. And just like this very serendipitous kind of interaction where you can run into people and exchange ideas. And I think it goes a long way to being able to iterate and conceptualize together quickly to remove, as you mentioned, kind of some of those gotchas later on down the line where– some the harmony of the form and the the necessity of the engineering don't clash at oh dark hundred right and and that's a super powerful process it reminds me of this technique that when i used to consult doing industrial design work uh for cpg companies we used to bring like a number of managers together in a room and we would do a training day you know half half of the first day would be a training day we'd bring them through some different techniques in brainstorming process. And one of the techniques is used a lot in a stand-up, not stand-up comedy, that's the improv comedy. And it's the yes and. And so this yes and technique when you're brainstorming or at the nascent conceptual development stages where regardless of disciplines, as you mentioned, you're building upon everyone's idea simultaneously to create the thing later. And it's this very unified process. it can be very kind of you know you jump back and forth quite a bit right but uh that ultimately will conclude with a better design in

SPEAKER_00: 30:40

the

SPEAKER_02: 30:41

long run

SPEAKER_00: 30:41

we used to we used to do these week-long design sprints where you would have you know a team of six people cross-discipline one from each one from each department

SPEAKER_02: 30:52

right

SPEAKER_00: 30:53

and the managers would come in midweek well they're coming at the start right they'd set the stage This is what the business wants to do. We want to get into this sector, right? And leave it fairly open. And then they would be like, they would come in, you know, midway through, and then they would come in at the end. And one of the levelers that I thought was really good was just the sketching of ideas on post-it notes with stick figures. And it's something as silly as that, right? Like where you force everybody to draw with the same level of resolution. there's often this, this barrier where I think when you, when you like, certainly for me, I like, I'm not, I'm an engineer that sketches, like I, my, my sketches are engineering led, right. They're more likely to be orthographic views. They're more likely to be like cross-section views of wall thicknesses and things like that. Right. So when I go and talk to a designer, someone that can sketch beautifully in perspective and, and, like it's a little intimidating right to go and sketch in front of somebody who you know is a better sketcher than you and so this this method of like everybody sketches their ideas in a simplistic way firstly helps break down those barriers it allows managers and non-technical people to to contribute ideas because they may have different insights right than than you as an engineer or a designer or um and It then allowed the managers to also get involved in that process and to have a discussion around all of the ideas without the best sketch appearing to come to the top because it's a good quality sketch or the loudest person coming to the top because they talk over everyone else or things like that. I like that idea of breaking down everybody's the same contribution to the process. You might come with different backgrounds, but you don't want any one to be too dominant, I don't think.

SPEAKER_02: 33:02

Yeah, absolutely. That's a really awesome technique. I like the fact that you brought up the point of drawing while you were brainstorming. besides kind of the stick figure kind of low fidelity drawing and having everyone be included in that process, I think there's something also that's something to be said about drawing as you're telling the story of what your idea is, because everyone can explain a thing. Like if I were to describe a product to you and have you draw it, but there's 20 other people in the room drawing that, what I'm describing, there's going to be 20 different drawings of what I've described. But when you're having this conversation with a group of people and you're all iterating and drawing the idea as you're talking through it, it also helps to kind of guide and direct as well. And I know it's a super kind of low... I was going to say low fidelity technique, but it's a super accessible technique to anyone to brainstorm, whether you're a designer, engineer, even a firmware guy or algorithms, for instance.

SPEAKER_00: 34:17

So just in the last year, I've started playing with the Oculus Quest. Yep. Love it. Using a software called Gravity Sketch, which is a design-focused... 3D sketching. I'll call it a 3D sketch, but it's actually a 3D form development tool in virtual reality. And I had a really interesting session with a VP where he had this idea and he wanted it. He couldn't quite articulate it properly. Not properly, but he couldn't articulate it to the people that he had talked to. where they would understand and translate his idea into what he wanted. Right. And so what we actually did was we did a live design session where it was really interesting. I had screencast what I could see inside my Oculus Quest to his screen across the other side of the country. And we had an hour where I would be designing in real time. in 3d and he would be describing and then i as he was describing i was creating form and then he was responding to that in real time and then i was adjusting the form and it was enabled because the tools had got to the point where real-time design had allowed for that and you know previously i would be like when i did my race car i was I was designing in 3D in SolidWorks and my customer was sat next to me. We were at a point where he was able to see and influence the design in real time. But coming into virtual reality has just taken that totally to another level. And it's enabled by the tools being so responsive and the speed at which you can generate form. is just incredible now. And I think that's really going to change how people go about exploring designs. We're even exploring taking those virtual designs and turning them into prototypes. Because you can

SPEAKER_02: 36:49

export out of Gravity Sketch.

SPEAKER_00: 36:50

Yeah, you're starting to get to the point where you can print directly from these kind of conceptual tools, which, again, 3D printing is traditionally been like okay you've got a generation cad and you know that might be an engineering based cad or it might be a like a design based cad but even even then like it was fairly laborious to um generate the 3d form to any any kind of like resolution where you'd want to print it but i think that's going to be changing and i think these kind of conceptual tools and how you You then can go straight to printing, which then brings back into the engineering performance side of testing. You can test those prototypes. It starts to merge it all together into a real-time product development workflow, which I hadn't really appreciated the power of that until you're sitting with a manager and you're able to do something in real-time. That was immensely valuable.

SPEAKER_02: 37:55

Yeah, that's the speed at which you can move nowadays with some of that is amazing. And that's a great example, Andre. I wonder if... So there's another facet in Gravity Sketch, which I know you're very familiar with, is that you can import and build around objects that you've modeled. And essentially, I'll say, remove the CAD goggles and see things in one-to-one relationships.

SPEAKER_00: 38:20

Yeah,

SPEAKER_02: 38:20

I think... I'm wondering if you could talk about that a little bit.

SPEAKER_00: 38:23

So, yeah, I... I actually did some development of some of the trucks that I've been working on in Gravity Sketch, and it's not really an automotive focused design tool, but I was basically bringing in the underlay components, so mechanical hardware as objects. And it's right at the very early stages of that being a workflow that Gravity Sketch has focused on. So it's a little bit clunky at the moment, I would say, but you can bring in geometry. It's kind of fixed at that point. You can't really interact with it too much other than placing it somewhere. But it does allow you to generate surfaces that can wrap it in a much more intuitive way than doing it on a flat CAD screen. I think the real power is going to come when you're doing that And I don't know if this will be through some, like I've really struggled to, sorry, I'm not making a lot of sense here. I think it's gonna be a lot better when it's an augmented reality design experience rather than fully virtual. I think that certainly when you're designing products that people are gonna touch and use and interact with, and they're gonna be in the home, like to be able to have that, if you imagine an augmented experience where you have a physics engine behind there right and there's there's you can see your environment you can have you know things in your hand that you that you're using in your environment i think like designing in that environment is going to be a lot better than um designing in a fully virtual environment um because and certainly things like the trucks where they're really big things that you're working on. But traditionally on a CAD screen, you shrink it all down and you wouldn't know whether you're working on a scale model or the full size thing. You have no idea. When you go to virtual reality, you still kind of have that, right? You don't work on it at one to one scale. You actually work on it like at the scale where you can see the radius or you can move the control points or whatever. So I think that's still a little bit of a step that needs to improve is like how do you design And certainly for consumer appliances and things where one-to-one scale is kind of the scale that you'd be drawing at anyway. Like Dyson, we always used to work with one-to-one scale cross-section views and all of the development would be done in cross-section. And you'd review physical drawings and we used to have to color code all of the wall thicknesses so that you could instantly see the stack up of different wall thicknesses and the air paths and all of the different components. I think that for those types of one to one scale products, then I think that an augmented experience is going to be a lot better for design. But there's no software that I've seen anyway that is allowing for design in a HoloLens type device. Everything seems to be virtual at the moment. And the HoloLens is more for like, okay, well, you've got your geometry and now you deploy it into augmented reality and you can inspect it or interact with it, but it's not really about creating it in that augmented space. And I think that will be really a good step once those sort of softwares start. And pass-through is coming with BigQuest and it's getting better and better. But I think that I would like to be able to do things like make a hand sculpted handle as a physical thing and then bring it into an augmented space and be able to sculpt surfaces around it that move when I move the handle. That would be a particularly compelling use case for an augmented reality design generation tool.

SPEAKER_02: 42:43

Yeah, no, I could totally envision that where you could literally wrap your hand around the thing you just made to see, you know, to check, not tolerances, but check the human factors relationship to that thing immediately, that immediate feedback.

UNKNOWN: 43:01

Yeah.

SPEAKER_02: 43:01

interesting yeah so you alluded to uh you briefly mentioned the uh the race car uh and i wanna i wanna just dig into this just for a second because i think it's super exciting is this the sky runner that you you worked on no that's the flying car is this the race car i'll just throw that away

SPEAKER_00: 43:20

the race car is one that i did um with uh a chap called adrian reynard who was who's you know a well-established race car uh designer constructor engineer uh you know had it had a company that had won the indy 500 and stuff like that so i i designed with him um he was a mentor figure for me and uh together we designed a uh a car from the ground up so all the components um you know we didn't design the engine but like We did a lot of engine modifications, but all of the chassis, suspension, and bodywork, and electrical, and things like that. Yeah, that was one that

SPEAKER_02: 44:05

I- Those files are available. I think I did a little research before we were chatting. It looks like some of those files are available still on, was it GrabCat, or what's the website?

UNKNOWN: 44:18

Yeah.

SPEAKER_00: 44:19

I had visions of thinking that open source could apply to hardware. And open source software is such a big movement and it's so immensely useful to everybody. I had visions of what would it be like if you just think about the intellectual property that's locked up in engineering drawings, like how to make stuff. And you think that none of that is available, right? It's all proprietary. For every product that exists, somebody's made a drawing, really. For every building, for every car, there's drawings and there's CAD for all of these products. And none of them are available to help other people learn how to design products, right? So this is something that I still can't believe, like, The amount of information that is lost to humanity really because how to make things to the right tolerance, to the right material, to the right shape, or how did you make that? It sits locked away in CAD servers and that information is not passed down to future generations of engineers. The only way for engineers and designers to create new products is to go through this learning process of trying to make something and then working out why you went wrong or like looking at, you know, you join a company and find out how the company has historically made things. And I wanted to, you know, provide some access to future engineers and designers to you know learn from what i had done right like i i spent a good few years of my life making this car and i created all of this cad and i had a full supply chain and i had all of these engineering drawings and 3d models and simulations and wind tunnel data and like you know suspension calculations everything and it was of no value just sitting with me right like it was no value to anybody else like it was only valuable if you were going to make more of those cars

SPEAKER_02: 46:39

right

SPEAKER_00: 46:40

and so i thought like it would be really good if i just open sourced it and i think it was the first open source race car um that anyone had done this is back in 2008 2009 and and so just published everything and let let people make it and you know people have been making it in sweden and russia and italy and uh the car and the molds are now we built a couple of cars i think they built another 10 in italy um the molds are now in new zealand so like and you know we've done a number of student projects where having full access to all of that data has enabled the students to then you know go and design new things or to simulate new things um on research projects on like improving the cooling or like even the other day i had some guy um on Facebook was like, oh, I downloaded your CAD, and I made some improvements to the wing position, and I did some simulation, and here are the files, and I think you could make it go faster next time, which is kind of cool.

SPEAKER_02: 47:45

I love that. Oh, my gosh.

SPEAKER_00: 47:51

I like the idea of, like, I've benefited so much from, like, open source software where you you have access to tools that like just speed up your workflow or things where features have been implemented by people and i wondered what it would be like if if if you could get to the point where any piece of hardware also had similar and i'm not talking about pcbs because like pcbs and like little cameras and you know raspberry pi's and stuff are all open source right but But from the hardware perspective, they're not big, complicated products, right? They're small devices. And I just wanted for people to be able to learn from like, okay, this is all of the things that go into making a complex thing like a car. And yeah, so it's all out there.

SPEAKER_02: 48:53

Yeah, now I know that– so as I mentioned, I did come across it in one of the CAD sharing platforms. But I'm wondering, do you have other– information baked into those models? Or how are you transferring some of the information that you're talking about that gets locked away to people when they download the parts? Is it just inherently embodied within each one of those models? What other features?

SPEAKER_00: 49:24

So there's a full 3D model. So there's multiple layers. Firstly, there's an assembly of all the parts, right? So you can You know, it's in Parasolid format. It's CAD neutral. You can download it and open it and you'll get a... It comes with the full assembly structure and all the part names and things like that. So, like, it's pretty good if you want to just explore around and you want to open up a sub-assembly. The way that you organize then the folder structure... I think I've just... Can you still hear me through the AirPods or is it through the...

SPEAKER_02: 50:01

That's okay.

SPEAKER_00: 50:02

Yeah,

SPEAKER_02: 50:02

I got

SPEAKER_00: 50:02

you. I think it's coming through the car now. That's all right. The way that you organize and you name your parts allows people to then dive from the 3D into the 2D engineering drawings, which have the same part naming convention, right? So you can open up any sub-assembly and then open up the... the drawing for that sub-assembly or the parts within that sub-assembly and then dive into the individual parts and see all of the dimensional data for that part. And I think that those then all obviously go to the bill of materials and the suppliers who could make those parts. And what I had envisaged was a way of okay, I've got this part, and you could then, because it's open and shareable, you can have quotes from the entire world's supply chain. Maybe a race car is not the right product because it's not high volume, but you could imagine a case where, let's use the vacuum cleaner as an example. I wanted to make a vacuum cleaner. I've designed it, and if I shared all of the CAD data Any supplier could quote on making that, right? It's not then protected. It doesn't require me to go to a supply chain and say, please, can you quote on this, which is how we do things today, right? We have a supply chain team. We have relationships with certain vendors and you send them some data and they will send you back a quote. what I want as a single person designer, I really struggled to stay on top of like, okay, I've sent it out to five suppliers for this one part to quote. And then you have to chase them for the quote. And then you have to like, like there was so much involvement of my time just trying to get quotes to make things. I thought, well, why, why am I trying to find suppliers to then give them business? Why aren't they trying to find me to, And, you know, quote, in order to win my business, right? I thought like it would be really good if the designs were shareable and you weren't protective about it. You could have quotes from anyone all around the world could come in and you could say, okay, I want to make this race car in America and I want to make it in New Zealand and I'm going to make it in the UK. And you could effectively have a supply chain that was an onshore, you know, supply chain for making the things that you want in the local market that you were in. And it would require not a lot of effort from my perspective, right, as the designer, because, you know, the quotes would come in and somebody who wanted to build that car could, like, review the suppliers that had quoted for quality or geography or price or delivery speed or whatever, right? Like, and they could select where they wanted to get it. But, like, I thought this would be a really powerful way of having hardware made locally for local markets with local supply chain without the overhead of having to stand each of those up individually.

SPEAKER_02: 53:26

Wow. Okay. Yeah, that paints a great picture of kind of where your head is at with some of that stuff. And I wonder... So to coin an Amazon leadership principle, I guess, think big, you definitely do a lot of that. And you probably rank pretty high when it comes to the quarterly and yearly reviews when people vote on those leadership principles. But I'm wondering, do you have any other techniques or processes for thinking big when it comes to engineering or design? Where do you find inspiration in that process?

SPEAKER_00: 54:02

I think the earliest, like one of my big concerns was that coming out of university that I was going to get stuck designing like a door handle for, you know, as part of a massive team. And that scared me, right? So like a lot of my early career was focused on trying to do everything just to get a bit of a, like I liked the idea of, knowing a little bit about a lot of things and it's it's a very that's actually like in retrospect it's quite a daunting um career path for young engineer to or a designer to think that they're gonna they're gonna defer getting a job as a junior engineer or designer in a big company right I think if you do go and get that job, right? Like it's financial security. It's like, okay, something on your CV. It's an intro to a job market. You get to learn from other people in a big company, right? There's a lot of advantages or perceived advantages. I didn't want to go in and design a little widget as part of a bigger team. And so that scared me. It felt like I was getting locked into becoming a specialist way too early. And so... I think my thinking big approach actually stems from that, where I said, okay, I'm just going to try and design the whole thing myself. And it may or may not be successful, but at least I will learn from each of the different aspects, right? I'll learn a little bit about wiring looms, or I'll learn a little bit about designing mold tooling or whatever it is. And when you have an appreciation for all the different areas that then sets you into a thinking big, like mindset right from the start. Right. So you, you train your brain to think about all of the aspects, which is really what design is all about is, is having empathy for all of the different disciplines. If I was to say, what is good design? It's, it's, knowing how to prioritize and balance the different competing KPIs or the different things that are important at any given time. And different products have different balances. It's like the old graphic equalizers, right? You've got all of these different knobs and levers that change how the product comes out at the end. And so as a designer, if you can't understand the importance of any one given lever, right? If you don't take an interest in all of the aspects, then you're likely to be narrow in your design approach such that your end result doesn't like fully deliver like where it could, right? Because you might not have thought too much about cost or you might not have thought about structural integrity or you might not have thought about like, how you're going to apply this texture in the tool, like whatever the decision point. And you mentioned it earlier, the weighting, how you weight the importance of each of these decisions is absolutely critical to the end result of the product design. And if you don't have experience in trying, at least trying or experience exposure to those different disciplines, you are not gonna be able to design a cohesive whole product. So I think the thinking big just comes from a desire to know a little bit about a lot of things and that then gives you a voice in all of those design decisions and that then appears like you're thinking big. But I just think of it as not wanting to think small.

SPEAKER_02: 58:05

What it sounds like you're talking about too is like, approach in different methods, right? Because each lens that you look through, whether it's the engineer or the supply chain or whatnot, there's a method to how they approach and succeed at the goal of their task, right? And it You know, when I was a young designer, even still in college, I was researching different methods and approaches to how to design. You know, it wasn't always just one way or one avenue or route. It's maybe not necessarily always just talking to a customer. Maybe it's the approach of making something more efficient by leveraging biomimicry or understanding lifecycle analysis so that you could reduce the overall environmental impact of a product from the beginning when you're designing it from its material selections, et cetera, et cetera. And I think having, you know, it's a skill though that it develops over time. As you mentioned, it's not something that, well, the naivety when you come out of college is that you're kind of, You want to try to do it all, or maybe you do want to specialize in something, but I think that there is some lessons to be learned that happen over time, and to not be afraid of just continuing to adapt and learn with the teams that you're involved

SPEAKER_00: 59:28

with. I would say that one of the most valuable things that I've done throughout my career is jump around from within different disciplines, and it gives you a really unique insight into the problems and the types of solution or the different approaches in those different industries, right? And those are immensely applicable in things that you didn't, you know, when you get into a new environment, you can apply a breadth of knowledge. I think that's really important for designers to expose themselves, which is why consultancies are so good as, you know, if you have different types of projects coming in. I think that's a really good fertile environment for young engineers and designers to expose themselves to different industries and different solutions and different problems. Yeah,

SPEAKER_02: 1:00:24

absolutely. Well, I don't want to cut our chat too short, but I want to be respectful of your time, Andre. I appreciate it. But I'm not going to let you get away... from this interview without talking about the flying car for a minute. We actually haven't chatted about it even since we've known each other and I'm curious if you can tell us a story of how that came to be.

SPEAKER_00: 1:00:49

So I was at Dyson and the guy that had hired me into Dyson, part of the interview panel, came to me one day and was like, I've got a friend and he's looking for somebody to design and build a flying car. And I thought of you and would you be interested in meeting him? So I went and met him at the pub and we had a great chat and turned out he had an investor in the US and he needed somebody to turn this idea into reality. And they had a year to do it. And I went home that evening and said to my wife that I think I'm going to go and make a flying car. And she looked at me. Well, she knows that I'm crazy. So there wasn't no surprises there. But basically gave up a well-respected design job at a great company. Dyson was awesome. And went to basically a startup where it was me and a fabricator. And we had to deliver a flying car in a year. So, um, yeah, don't be afraid to do crazy because when you push yourself outside of what you thought you could do, like, you know, you find ways to invariably you find ways to do most of it. Like there are things that I would do better, but the flying car was, was definitely a better, uh, product. Like it, it was a lot more, It was after I'd done the race car, so I had a lot to leverage in terms of how I approached it. The tools that I would use, the supply chain, I had a lot of connections in the industry that could make things for me quickly. I had a lot of CAD to even as building blocks. And so it was a lot more efficient as a process. Things went straight from... you know, the computer to us, you know, go and digitally laser cut these tubes. And, you know, a week later we had a chassis like that was a lot faster. You know, the race car was a lot more handmade, even though we had CAD and designs, it was like, okay, you've got to set this tube up in and we're going to, you know, chop through it with a whole hole saw. And like, it was a lot more, um, yeah, tools had moved on and digital fabrication was a lot better as well. So, um, Yeah, it was really good. And then we shipped it over to Florida in a box and I assembled it. I assembled it in... We launched it at a boat show, a super yacht show in Florida. And it arrived in the parking lot in a shipping container in pieces. And I was there with my... tool bag right like in the parking lot assembling a car

SPEAKER_02: 1:03:49

putting this thing together

SPEAKER_00: 1:03:51

and then we put it on the dock next to these super yachts and it was great it was um it was a fantastic experience and uh

SPEAKER_02: 1:03:59

Yeah. That's awesome. Holy cow. Well, people can check that out in the show notes. I'll grab a link and obviously attach any other handles and such that you have, Andre. Where can people go to learn more about you, follow you, keep in touch?

SPEAKER_00: 1:04:19

I don't know. I've stayed away from having too much of a personal profile out there. I expect LinkedIn is probably my... If people want to chat, I'm more active on LinkedIn than I am on anything else. I really should do a little

SPEAKER_02: 1:04:39

bit

SPEAKER_00: 1:04:39

better on having more of a design portfolio type, explaining how I've gone about creating certain products. I'm particularly proud of the Dyson robot I did all of the mechanical design and the layout and there's some fantastic things that have stayed through the subsequent generations of Dyson Robot where they've kept the same layout. That makes me proud. But yeah, I should really do more of a portfolio page that sort of explains the design thinking behind why does this product that I worked on look like that and what reasons and what innovation enabled it. you know, explain a little bit more about the process that I've gone through.

SPEAKER_02: 1:05:27

Yeah, no, I mean, that's why we're interviewing you today, Andre. I mean, we're just going to unravel it here.

SPEAKER_00: 1:05:33

That was a really, if you've got two more minutes, that was a really interesting.

SPEAKER_02: 1:05:37

Yeah, yeah, absolutely.

SPEAKER_00: 1:05:38

So when I started on the Dyson robot, like all robots, home robots, home vacuum cleaners have this bumper, right? That goes around the front of the product. And, you know, they drive around and they have an array of sensors and things and, Sometimes the sensors miss objects, right? And so they put this bumper on there so that if they hit into anything, they can have a physical microswitch that goes off and says you've hit into something, right? And you might have more than one microswitch and you might know which direction you need to back away from that thing. But in general, that makes all the robots look the same, right? They've all got this big wraparound bumper on the front and... From a design perspective, they're very vanilla. They're hockey puck. Everything looks the same. Dyson have this technology that they've always had on all their vacuum cleaners. It's all cyclones. Cyclones are round. All robot vacuum cleaners have square, like they're basically sweepers with a little bit of suction assist and they sweep into a square little tray and you'd lift the tray out from behind some cover or something like this, right? So they're all the same. All the other vacuum cleaners are the same. Dyson's like known for its cyclones. In all the other products, the cyclone is front and center, right? So how do you enable that on a robot that needs to have a bumper This is one of the cases where I had to really fight hard in order to take a risk. And the idea was one that had come to me. I'd seen, I think it was Opinion Farina car design, where they had taken inspiration from an egg. Now, this sounds like a totally... roundabout-the-houses way to get to a solution for robot vacuum cleaners, but this is exactly my point about immerse yourself in all types of solutions to problems in different industries. An egg has an important bit in the middle, which is the yolk, and it's suspended in a secondary fluid medium, which is the white. part around the yolk and that's then protected with a hard shell around the outside right so if if an egg hits something the yolk moves within the within the egg right there's no bumper on it on an eggshell right yes yeah of course so the the the movement they give you have to have movement against a spring in and a damper in order to dissipate the energy right so without movement your peak accelerations are really high, right? Because you have no time to slow down. So in an egg, that's how nature solved this problem of how to protect something. Now, Pininfarina thought, like, hey, that's a great idea. Why do cars crumple? Eggs don't crumple, right? Like, if an egg crumples, then it's failed, right? So... In nature, they solved this problem by having the movement happen internally. Pino Farina thought, hey, that's a great idea for a car. Why don't we have the seats on an internal sled, right? And have the car shell be really rigid so that if you hit into something, the seats give, right? And the whole thing moves. And that kind of happens a little bit with like steering wheels and things like that. But they took it to another level and they had the whole seats move on a sled just like the yolk of an egg. I had seen this having been interested in cars and interesting ideas and biomimicry and stuff like that. And I thought, well, what would it be like if you had your robot have a rigid outer shell and then have a bunch of the internal components like the drive systems and the batteries and the air pathways and stuff move inside the outer shell? So the outer shell becomes your bumper, effectively, but it's all the way around the outside. And so when you do that, it means your design can get cleaned up because you don't have to have a movable bumper and part gaps and all of this messiness that happens with current bumpers. They have to allow like 10 millimeters. So anyway, I advocated to totally remove the bumper and turn the entire outside of the robot into a rigid outer shell and then have the insides move and then contact microswitches so that you could slow down the inside and there were springs and microswitches and things like that. And so that became a patentable feature. The other thing that it allowed was the entire cyclone to be part of the rigid outer body right so

SPEAKER_02: 1:10:50

okay

SPEAKER_00: 1:10:50

that allowed the cyclone to be where the bumper used to be which totally changed the industrial design of the product and allowed for this really cool well i think it's cool um it was also in the museum which which yeah yeah um but it but if you look at it plan view it's a it's it's a circle with another overlapping circle for the cyclone that sort of intersects the wider outer circle. It was a design that was copied by some other large Japanese and Korean electronics companies, which made me chuckle. I went to CES one year and I'm not going to name names, but there was a large, they make TVs and other consumer goods, but they had basically totally ripped off the design and that there was nothing more pleasing as a designer to have um but it has this this idea of having the the part that the customer needs to take off the product right you don't want to hide it behind a mechanical system that's that is not serving um The ease of access, right? To empty the bin and to see the technology that is different was really important. To be able to show it at the front of the product to say, our product is different because it has a cyclone. Here is the cyclone. It's easy to get to. You don't have to open a trap door to get the dirt out and stuff like that. All of those things were really good reasons. I had to advocate really hard because... the risk was that you would not be able to sense the bump quickly enough and you would bump into things and damage things right so we had a lot of um proving out of like can you can you slow down quickly enough can you make the bump soft enough so that you can not damage furniture and other things that you might bump into um and that was quite a hard sell um you know, to a robotics team that wanted to rely on a bump signal,

SPEAKER_02: 1:13:07

it

SPEAKER_00: 1:13:08

ended up creating a really unique industrial design that's quite clean and pure. You know, it's two circles. And then, like, fortuitously, there was also a camera that was in the middle. So you ended up with a large outer circle a smaller overlapping circle that was approximately half the diameter of the big circle and then another smaller overlapping circle for the camera that was like again it was like this cascading sizes of circles that all over and it was just really really nice afterwards and I can remember sketching it just like three circles and it was like and that has persisted and I'm really proud of like That as an example of taking a solution from a totally different industry, from nature, and applying it in a non-obvious way, and the result being something that is... unique from a design perspective but has some real reason to to to be there from a customer uh point of view i i like that's that's been

SPEAKER_02: 1:14:15

wow

SPEAKER_00: 1:14:16

a really nice project to be involved with even though i was only involved in it for um you know the very early phases of like creating the you know p1s and p2 prototypes laying out all the components and um There's a lot of other people that took that product and solved all of the problems that I created and tried to get it to market.

SPEAKER_02: 1:14:40

Wow. From egg to race car to home robot vacuum cleaner. I tell you, it's a perfect example of biomimicry, but also, as you mentioned, a perfect example of exploration outside of your realm, outside of the categories.

SPEAKER_00: 1:14:58

We made a clear outer shell. And then you can see all of the components inside and the springs and stuff. And we made a little

SPEAKER_02: 1:15:06

deceleration.

SPEAKER_00: 1:15:07

Yeah, we made a radio control so you could drive it into things. And you can see it hit into things. And you can see all the insides move. And it's just like, oh, yeah, that's so satisfying.

SPEAKER_02: 1:15:17

Oh, that's perfect. That's awesome. Well, how about one more question here, Andre? I'm curious to know what excites you right now in any realm? It could be. methods it could be design it could be engineering it could be new technology i mean i'll let you go with it

SPEAKER_00: 1:15:34

i i touched on it earlier i'm super excited about augmented um designing in an augmented way um i i think that those tools are coming on uh like i really i i follow quite closely like the you know the photogrammetry and the oh actually the thing that really is that i haven't dived into this

SPEAKER_01: 1:15:55

yes

SPEAKER_00: 1:15:55

It's called procedural design, and it's used a lot in people that make game environments. So it's in software like Blender, right, where you can... It's kind of like you've got these mathematical models and you've got nodes that you can connect different aspects, right? You've got shaders, you've got geometry, you've got rendering, you've got all of that, right? And you can basically... tune your sliders and tune your math models and then connect different inputs to outputs. And I haven't even started doing it. But the things that I see people do when they are procedurally creating game environments or they're making buildings or they're making pipe layouts or they're making... And you just see these things magically grow from the definition within the software. I'm really interested in could that be applied to physical things that you want to make, right? At the moment, they're being applied to create these fantastically rendered images meshes and geometries but i'm interested in could that be applied to physical product designs

SPEAKER_02: 1:17:09

well so this you're talking about so this is uh something i use quite a bit in grasshopper in rhino have you yeah

SPEAKER_00: 1:17:16

i've seen parametric modeling yeah so i'm so that's an application to like okay i've got a repeating pattern of holes or like little valleys or like whatever right but That's the first step into procedural product design. I'm interested in, could you design the air paths for a vacuum cleaner procedurally? If you took a complex product that has all of these different design elements, some of which have engineering and mathematical models that define them or define what's good, Could you layer all that together into something and spit out a design at the end that could be printed? I'm really interested in that intersection of algorithmic design, mathematical, like all the engineering can be represented in a physics model, right? So this intersection of game, render, simulation, mathematics, and physics, and then the experience of designing products. like there's a whole load of experience that goes into say, well, this is good. This is not so good. Right. Like if you could embody all of that into a procedural algorithm that, that can help you design products, I think that that's going to like rapidly accelerate, like, product design, I think.

SPEAKER_02: 1:18:41

Yeah. I share very much your energy and passion for this approach because I've been toying recently with, in Grasshopper, you can create equations and relationships between different forms and lines and splines and et cetera, et cetera. And what that allows you to do is you can create a gross form. Let's say it's like an arc or a shell of some sort, and it's defined by a particular profiles and lofts, which are defined by the instances of the nodes of each of those lines, but you can put a slider or some sort of metric to each one of those physical features and move things and scale them and create like infinitely adjustable outcomes or final results on the fly, live, without having to model it, and you just bake each one of those, and then you have 100 permutations of something that is all relationship and constraint driven.

SPEAKER_00: 1:19:40

So if you think about a product manufacturing model where you're printing those things. you're going to get to the point where people can customize their own products and have them printed and shipped to them. And I'm talking about vacuum cleaners or things like that, like consumer products. And we haven't got there yet, but at Shark, we were printing production run with hundreds of prototypes and having them assembled on the production line and still in the prototype phase, but having them properly assembled as if they were real tooled parts. I think that could be a future as printing gets better and the materials get better. I think that could be really interesting. And then to marry that with algorithmic design, I think it was really exciting for me.

SPEAKER_02: 1:20:39

It's going to be wild. It's going to be wild. No, I'd be excited too. to mess around with you on a project or something like that, creating some sort of ecology of procedural things that, you know, as you move, one thing moves around and then this kind of embodies something a little more. And, you know, whether it's like how the thing is made or how it's put together or whatever. Well, we've touched on a lot of different stuff, Andre. Again, thank you so much for your time today. It's been a, yeah, it's been super, super fun. I'll let you get back to charging your Tesla. oh nice very

SPEAKER_00: 1:21:15

good i'm incurring i'm incurring like charging fees now for sitting in the charging

SPEAKER_02: 1:21:23

oh god all right

SPEAKER_00: 1:21:24

well

SPEAKER_02: 1:21:25

yeah you're that guy right now all right cool andre thanks so much we'll be in touch again soon uh just uh i'd love to catch up again maybe catch on the ski slopes out here in new england

SPEAKER_00: 1:21:35

yeah

SPEAKER_02: 1:21:35

sounds good all right cheers see andre thanks a lot bye