0:00 All right, sorry about that, hello. So my name is Puneet Jhaveri, I’m an applications engineer at Carbon. My background is: I was originally a premed student, then decide, through, like, taking some classes in architecture, industrial design, that I kind of wanted to transition, and I found mechanical engineering to be a good mix of both of those things, so I love trying to apply kind of that background to mechanical engineering to design custom products.

0:34 And I am Sean Fier, head of design engineering at toolkit. I also have a background in mechanical engineering, I started my career at uskin, one of the first 3D, fully 3D printed knee brace that was developed out of Canada, but also, as I moved with tokit, I got involved in a lot of the elmet project, especially coming out playing football in college. I’ve, I’ve known a lot of people who had trouble with their helmets and a lot of injuries, so those are project that are close to my so really happy to talk about those today.

1:07 So yeah, so what we are going to go over is actually how we both worked on, with CCM, to bring their custom helmet to scale.

1:17 So a little bit about our background, is Carbon: so we have industrial hardware, design software, and automation software, and engineering gr materials that all kind of work together to build unique products in different spaces, from consumer, to oral health, yeah, to protective padding, things like that. Here are some of our products, so you can see that, you know, we have production products at scale, can see saddles, gloves, backpacks, so performance, protective, and we’ve actually to date have produced four million lattice products, produced on printers around the world.

2:03 This is a snapshot of our design software, design engine, so in design engine you can do complex latticing, where you can tune for different mechanical needs, you can also bring in custom inputs and connect to an API to get data in and pull data out.

2:25 So our vision at Carbon, you know, we’re really proud of all the the products that we brought to scale, but we want to start moving more towards the mass customization, to bring it out to to the masses, to bring the products to scale, and build tools that are repeatable across different products and industries.

2:46 So diving back, kind of bring it toward CCM, to kind of walk through it, almost as a small case study, at least our sections. To be able to do that, you need innovations in computational design tools, so you need to be able to bring in the custom inputs, you need to be able to automate, you need to get data and understand what’s going on during the process. You need high performance materials that make sense economically for the customer but then also provide lightweight needs and fast print times.

3:17 You need a true onetoone product pipeline, right, to be able to understand, like, hey, what’s that scan data coming in, how do you, how do you use that data, and that’s what the experts at toolkit 3D do. And a new digital means of manufacturing, right, you need to trace that product through the system, you need to pack it, you need to put on a build, make it efficient to print.

3:40 So on a high level, I’m going to go through kind of the design side of that CCM helmet. So where do you usually start, right, and I think some of you might have seen this actually in slice Labs presentation of a great place to start is understanding what lattice to use.

3:56 So we usually, as an applications engineer, you guys provide us with the material, we characterize it, and then we can use our thousands of lattice library elements that we have to kind of narrow down what lce might work for you, and this is actually available to all our design engine users, that you can go through this. So it’s a mix of experimental data on luses and also simulation data, and then we fill in the gaps by interpolating from from both of those points, and we’ve also recently plugged in kind of commercial available foams in there as well, so if you have something that you’re already working with, you can kind of understand the compression curve and find what sort of lattice works best for that.

4:40 Hand inand, when we’re trying to figure out what lattice works best, or at least narrow down the scope of it, we’re figuring out what material also works best. So for this project it was a little bit easy, because we have been developing epu 45, which is our premier dampening elastomer, strain rate sensitive, so it was, it was a perfect fit for this project. With this elaser we’ve actually been able to improve printability from our previous U 42, and we cut the print time, print time, down by over 70%.

5:18 Once you kind of understand what material and what lattice you’re going to use, you know, we, we have to start tuning that structure. So, within design engine, we’ve been able to use our advanced strut selectors, and I’ll give you one example here, is we actually get the head form of the player, and we can understand kind of what’s normal to that head form, to understand where you can and you cannot change strut diameter, right, where your performance struts are, and where you have other struts that you might be able to cut down in terms of weight, so we use that and and reduce struts in the non-normal direction to bring the weight down for a player.

6:04 And the other thing, as I mentioned, is automation, right? So in this process, as you can imagine, you have, you’re working through a bunch of iterations, you have different scans coming in from from the customer of different players that they want to test and iterate through, you need a quick way to do that versus starting a new project each time. Within design engine you can actually just have a drop down at the top, you can tokenize your inputs, I don’t even know if you could read any of that, but you can just select the input, give it a token name, and download a script.

6:31 So for the CCM project I’ve actually used that, and and built on that, like, an own python script on my desktop, using this as the starting point to be able to just run through helmets and helmets, and these helmets get very complex, right? You have, from toolkit 3D who processes scan, we have seven different inputs coming in for a goalie helmet. The go goalie helmet has 14 different pads and comes in three different sizes, so you need kind of a way to reduce that, to reduce the design time and iteration time and efficiency.

7:02 And then you can also imagine when you actually build this into actual production workflow the complexity that I have to build in there, and that’s why oops, yeah, that’s why we rely on people like toolkit 3D, who’s our premier integration partner. So we kind of handle the design, latus generation, and the print and production side of things, and they interface with us through our API, to to bring everything through for for the customer.

7:32 So yeah, I’ll pass it off to Shan F. Perfect. So first of all, I’ll just take us one step back to kind of show the initial input at to that we had, a toolkit.

So in that project we were both brought independently, carbon and toolkit, by CCM. Basically, what we were, were both pieces of the big puzzle that we want, that we needed to solve, on their side was designing the pads, making sure they’re printable and they’re gonna follow the guidelines for the testing. On our side was how can we follow some guidelines and make sure that every single time we receive a scan of any player, we’re going to be able to generate the volumes that are adapted to their ad, and they, and we’re going to stay within the constraint for morphing, so it can be either maximum thickness and minimum thickness to be within test standard, or it can be just printability assessment, so making sure that some surface are flat, so they’re going to stick properly, we’re not going to have overhang, or make sure that we don’t need any support, so we can optimize as much as possible the print.

8:38 So that came as our design engine. So what we input initially is a raw scan, as you can see the the player right now is wearing kind of like those at that are there to compress the hair, then we’re going to run it through our Ai and orientation. So it’s an AI that we’re using, 83 line landmark, that are able to be recognized, because when you get a scan, you have no CL, the computer has no clue what he’s dealing with. So the first thing is understanding, making sure for it’s ahead, and also where the where is it in space, so that AI allowing us to make sure it’s ahead, and also orienting every time to make sure that, you know, the nose is at the same play, pointing forward, the head is upward, and things like that.

9:23 I’m not going to fall into details for the two other step, because those are the propriety parts of our customer, but afterwards, what we do is we are able to remap The Edge completely. So we take a scan, we’re going to be able to take my head, we’re going to be able to take pun’s head, and we’re going to be able to remap perfectly every single point on that head, so every time your nose is going to be exactly the same point, your ears are always going to be exactly at the same point. So that’s going to allow us to do dir rectification easily, so for certain portion there’s air, so if you want to remove the bumps around there, if you want, if you want to do some compression, if you wanna, so any type of rectification that you need that are product specific, that remapping is going to allow us to replicate this automatically every single time.

10:11 And then we move into the part morphing. The part morphing is twofold, like Ping mentioned, every element comes in three sizes, six different position. So the elmet is over here for people, okay, there, stickers, I’m not going to play too much, but basically you can open up some flips, and there’s six position, front to back, that can allow you to extend the helmet.

10:35 So basically what we’re able to do through our rectification and measurements, the first thing, we’re going to fit into the right shell, so basically hopefully minimizing the pad thicknesses to minimize weight, and then we’re going to following the match to the right cell, cell size, sorry, we’re going to be able to morph the pads to properly fit your head. The pads are compr of, of course, the volume, but also their skins for sweat and comfort that are morph at the same time, and then that are pushed to the carbon design engine.

11:07 This is all fun and everything, but we soon learned, all three of us, that was only one piece of the puzzle, because now we add every, not necessarily one piece of the puzzle, we now add all of the pieces, but we needed to put them back together into the final picture, because we have now our product, we have a scan, we have scan processing, we have morphing, and we’re able to lce it and print it. But what happens when we move that product into production? We’re talking about a complex product, we’re talking about multiple design engine, we’re talking about multiple steps.

11:41 How was this being done initially? Of course we were using email and SharePoint. How can we scale production for hundreds and hundreds and hundreds, and moving to thousand of players in the future? That was one of the biggest paino that we figured out with with CCM, and this is where the order management came in from toolkit.

12:04 So basically, it’s a play on word, would have thought, but toolkit is basically an open toolbox, we’re there to bridge the gap between different pieces, if you don’t have those pieces, we’re going to help you build those pieces, or you can bring your own. In this case, we have the carbon design engine, we have the carbon platform for printing, and we have some scanning app, so basically we’re bringing some of on our own tools, but also connecting with the already existing tool, to remove human interaction and delays that are non necessary.

12:33 We are also able to keep traceability through the whole production, as you can see, this is just a quick view on the right side of our platform, but you have kpis on ordering, so basically you’re able to track how many orders are going through the platform every week, months, year, you’re also able to track every single product, at what step it is, as well during the production line.

12:59 So if we go back to our, to to our flow, basically we’ve been able to figure out the first four steps, so up to print and production, but right now the next step, what we’re trying to figure out, is going to be production. So right now we assess that carbon’s API are easy to access and interact, we can even connect with their Fleet control, allowing us to see when parts are being queed, what, when parts are being printed, and even know, know what’s the time left on that specific print.

13:30 Bringing all that back to the OMS platform, so basically the CCM, or any partners into the production line, including printing partners, have a single place of through in the OMS. So no need to move from engine to engine, no need to move to software to software, everything is streamlined and available on a singal platform.

13:53 We can even input some QA steps, so basically there’s a viewer on the platform, so between the morphing, between, once you have the lattice part, you can check in real time and make sure that everything is following the process before you send in, sended it to the next step. So, of course, we want to print functional parts.

14:16 Then pushing even further, carbon and continue to develop new, new tools, like the carbon Automation and PR, print preparation software. So the next step is going to be taking every single one of those element that are being product produced automatically, creating some nest file, net builds, that can be put in a queue automatically.

14:38 We’re also planning through some, the network, the printer network, that they have currently, to be able to distribute the order according to different characteristic that could be geography. So basically, if you’re in Europe and you want to order an elet, you’re not going to print it in North America, you’re going to want to print it somewhere closer to home, because why would you ship the parts across the pond, that makes no sense.

15:04 And also ensuring that, that through your production network, you have even, sorry, you, you ensure that you have printer availability as well. Once, once you’re creating some orders, we are, we also right now have taken delivery of old L1 printer, and this is multiple folds. First of all, is it gives us access to the architecture, the software, allowing us to understand a little bit better, play with it, and and better build tools that can create synergies.

15:40 But also, we have a lot of customers that are coming to us asking, oh, we want to develop some product, we’re going to do some consultancy, we’re going to use the design engine, we’re going to build some parts and do some quick turnaround on prototyping, to hopefully show them the potential of additive manufactur and carbon technology.

16:01 Right now we’re in the Fulfillment phase, so CCM is still in the work story, it’s already a success story, but it can be even a better success story. Last year during the playoff, 25% of the players was using the super tax, super tax helmet, but it was only available to Pros. So basically, it was sales rep going into Pro locker rooms, getting them scans.

16:26 Right now, CCM is releasing those scanners into shops, your sport, sorry, I’m from Canada, so I might name drop some shops that you have no clue, but those are the ones that I’m used to, your sports, your SP. And so basically, any sports shop that you go into, they’re going to have they’re going to be trained, they’re going to have a, the scan app, and and basically everyone’s going to be able to order a CCM custom helmet.

Basically, the only Grail of what we’re trying to achieve with them is within order to deliver to customer, to, we want them to have the helmet within 10 days. With what we’ve already be built together, we’ve already been able to remove between three and four days already in that workflow, so still a work in progress, but a lot of Promise showing.

17:24 I really need that coffee, I think, I think I see some people sleeping too, I think some, some other people do too. So yeah, so we, we are growing, growing our customization Solutions, you know, with prosthetics, cushioning, we have things out there, but we are looking to go further. So yeah, if, if you guys have products that can grow with us, and even thinking of it in your prototype space, iterating through products, if we can help with our design engine or the full final production scale option, come talk to us. Would you have some samples to push on? We have helmets to try on, the boots right there. So yeah, happy to talk to you guys. Thank you for your time.