0:00 So, we’re Rapid Liquid Print. And our presentation today is called Design for Viscosity, Not Gravity, where we’re going to be highlighting not only our technology and our company, but also our collaboration with Coparini and our production run with them. So, what is rapid liquid print? It is the process of extruding a silicone into a gel suspension medium, but that enables what we call gravity-free manufacturing. So, gravity-free manufacturing is comprised of three key components, which is the true RTV platinum cure silicone, which is a production grade ready silicone, and it’s all extruded into the gel suspension medium, which allows us to go completely supportless and do parts really, really quickly.

0:46 And it’s all built around our topology first tool pathing software which is completely custom to our process parameters and understands every shape in its context for our gravity-free system. So we use this to make completely isotropic parts with high speed and production grade surface qualities. So how is it done? It’s basically you print it, you pull it and you wash it. So there’s no additional post-processing about the system and it’s it’s pretty fast.

1:17 So how do we get here? Seems simple but it all started at the MIT self assembly lab in the MIT architecture department a few years ago. This lab was focused on self self assembling technologies and our early projects were on shoes for instance. Here we have all manually drawn tool paths producing a native shoe with the company native and we our early projects were mainly focused on design and fashion and things like that.

1:48 So here we have a liquid printed products page with Kristoff Uberon and Marcela Colo which was exhibited at Design Miami in 2017. And additionally we did projects for a variety of household items with the Patrick Parish studio which was acquired by the MoMA and displayed with the light cycles exhibition recently. So those early projects in design and fashion were quite interesting but what ended up happening was that we wanted to go into more production ready applications.

2:21 So because we were doing everything with this true RTV silicone, we felt like it had real world applications. So we got interest from BMW to start doing applications in actuators and inflatable seat cushions which afforded us the ability to spin out into a startup in 2021. So here are founders in the original garage of our CEO and the machine that we bought off of MIT in 2021 and and when we moved to Charletown, Massachusetts, which is right across the bridge from MIT if anyone knows the Boston area.

2:54 So those early BMW projects were mainly about monomaterial seating. Here’s one where we have an inflatable lattice built inside of a custom cushion for early stage concept cars which kind of you know culminated with this project with design works where we made this visionary material seat which our contribution to that was the lattice spine which is a multiaterial part so multiolor and multi- durometer part in the same print which ended up winning a sustainability award from Altier but but Our bread and butter has always been orthotics and prosthetics.

So, everything is mass customized. Everything, needs to be human first contact. So, our softness of our silicone is really great for these processes. And we end up starting out working with Bionic Skins, which is another MIT startup by Hugh Hair, selling them an early pre-BA machine that allowed them to print the proc the parts that they weren’t able to make in any other method. Additionally working with projects like Namaste with Jeff Arenstone and at the time we were delivering software to them as well.

4:02 So software that allowed them to make the parts with the machinery that we sold them directly in-house without understanding any of our unique process parameters and just clicking print. So, you know, people might recognize that this is like an early human UI like skinning of, you know, Rhino and things like that. Just allowing people to print what they need to print in a very simple geometry constrained manner.

4:24 During this time, we did hardware iterations as well, going from our alpha machine on the left, you know, all 8020 to beta machine, which is what we were selling. And we got to a point where we felt like we could start picking up more complex, bigger production ready applications. So that’s what the main meat of the conversation is going to be and I’m going to hand it to Kimble here, our head of production.

4:49 Right. So I’ll introduce the sort of Copernie section of our presentation. That’s Hamilton by the way. He works on the software at RLP. He gave me the introduction but not himself. So that was sort of the company history that brought us to the Capern collaboration. And just a quick Capernney introduction. They’re a Parisian fashion brand with sort of their own unique design language, but they’re also interested in what they call tech fashion.

5:17 So they always are looking to experiment with sustainable and innovative materials and they often pursue future forward methods of production. So, probably many of you have seen the sort of famous piece that they did in 2023 where they actually created a dress midfashion show by applying a liquid fiber material directly on top of a model and then shape the dress live in front of of the audience.

5:44 And they’re also known for their swipe bag series, which is essentially a series of variations on the same bag profile, but they’ve always made it in different materials. And so that’s where we also came into the collaboration with them. So we collaborated on the spring summer collection of 2025, which was also a special show between Copernney and Disney as it was being held at Disneyland Paris. So we enjoyed that.

6:13 Capernie was interested in working with us at RLP for printing a bag, but they were also specifically interested in our technology because of the way we print in a water-based gel, which for them was a perfect connection back to Disney. And therefore, they called our bag version the aerial swipe bag because it’s supposed to connect to the mermaid and water sort of feeling. So, that was a first for us, but we appreciated that, too.

6:40 This is the original swipe geometry. When it’s made out of leather. So, we knew from the start that this is the the rough shape that we would be working with. They’ve done projects where they’ve done it in glass and crystal and even they’ve made it out of a NASA developed silica aerogel. So, we weren’t the first ones to do something outside of traditional leather. But of course, our version was in silicone.

7:07 So this was the sort of starting point that we had gotten from the Capernney team which was again that same shape but they had gone for sort of like a smoother geometry in a way that we thought would work well for sort of the RLP silicone version. And there was essentially a set of design requirements from the designers as we worked on the bag project together. First the profile of course needed to belong to the swipe geometry.

7:37 So there was limitations to how much we could change that. The bag the swipe bag is specifically sized to hold a smartphone. So ours still had to do that same thing. It was supposed to be flexible because of course it is made out of a flexible silicone. But the bag also needed to be rigid enough so that when items are placed inside of it that iconic profile doesn’t bulge or change.

8:01 We also had a challenge of sort of trying to make the base as flat as possible so that it could potentially like rest and sit upright when it’s sort of set on something. And then of course we just needed to match multiple colors that Coperni designed for their show and then also match the Capernie branding on this bag too. So to start, we went for what we thought was going to be the easiest part of the bag, which was trying to make the flat bottom so that it could sit at rest.

8:32 But actually became one of the most difficult. We printed several iterations of the bag in our silicone to see sort of what geometries would stand on their own. And you can see this in the glass versions as well. The bag starts to like droop a lot to get it in the glass version to sit. Which was actually a really similar sort of situation on on our bag.

8:53 However, after multiple iterations and going through this with the Capernney team, we decided that actually we’re going to go with much subtler approach, which is the sort of far right version. Where it wouldn’t necessarily sit on its own, but it would still just have a resting point when you lean it up against something. So, we really wanted to preserve that profile of the bag as much as possible.

9:17 And then one of the more aspect one of the more challenging aspects of the bag geometry which we knew was going to be challenging from the start is working on how the opening would work in the bag. Sometimes people see the bag and think there is no opening but there is indeed an opening. It’s a bag you can get inside of it. So Caperni wanted to do sort of the most minimal slit on this bag.

9:38 We had tried different versions of the opening but went back to sort of the most minimal slit is what we wanted for the final. So this was always going to be like a a a particular a particular focus area on this geometry for us for a few reasons. The way that we tool path parts we have a few few ways to do it but generally speaking we always try and do a non planer continuous tool path.

10:07 That’s what always gives us the most isotropy and the fastest print times and the smoothest surface finish in our parts. So we sort of have some guidelines developed on our specific machine that we know we have to sort of adhere to to get quality prints. So when we’re making these non-planer tool paths, we basically set a constraint that we know we can travel in sort of a direction in 3D, but it has to be in sort of roughly a 45 degree range off a flat XY plane.

10:41 And that’s traveling down as well as traveling up. So of course like pre-planning, we knew that also we were going to have to probably break this geometry up into a few sequence tool paths, which I’ll show, in a minute. But we also knew that we need to have sort of that slit opening be the ending boundary of the first tool path. So you can see here if we draw the the needle extruder on top of that opening, it does become very steep there.

11:12 So again, just to show that in 2D profile, the length of the slit is that purple line. And this was also a pretty strict constraint because we knew that it has to be this length to actually get inside the bag and do things like put a smartphone inside of it. So we also knew we somehow had to achieve this by also keeping that iconic profile. But if we zoom in sort of further and break it down and examine some of the points of that hypothetical boundary, if we draw sort of a tangent line at several points across it, you can see we’re going from like a flat zero all the way up to like a 75° angle steep, which is probably way too steep for us to get a nice finish.

11:54 And then the other challenge was again we couldn’t just shorten it to like where it says like 47 degrees because that would make the opening way too small. So the strategy that we came up with the team was to gently modify the profile of the bag and essentially bring that opening area down into a more forgiving angle. And so we went through several several iterations with the capernic team to do this cuz obviously that’s very important to them.

12:20 And so we were trying to make it perfectly calibrated so that the length is still the appropriate length of the opening. But again we’re still like not destroying the iconic profile and so on and so forth. So quickly just in 3D that meant that we had so developed a script to basically just morph the mesh model to that profile that we had drawn. So there’s just a fall-off distance on that sort of profile and then we essentially just moved vertices on that mesh into that appropriate shape for printing that better boundary.

12:55 And then that would be the updated boundary there or the updated geometry, excuse me, there. The thickness strategy was also something that we needed to develop with the team. So originally Capernie was just asking for a solid handle. But we knew we could probably work with that. We also needed to test basically at what point and what thickness does our shore a50 silicone not bulge with stuff inside of the bag.

13:24 And we had found that actually you can do basically a four to five mm thickness throughout the body to the bag. And that was enough to have enough rigidity. And so we at the same at the same level, we did multiple tests where we knew we needed the handle to be rigid, but it didn’t need to be solid. We we changed it to sort of a gradual change in thickness to about a 6 to 7 mm range, and that was enough to keep the handle rigid and stiff.

13:47 And so we approve we improved basically the print time as well as reduced the amount of materials that was needed for the same print. And then just going back to mapping the tool path on this geometry. If we imagine trying to draw a continuous spiral across this geometry, you get to a point of course where the handle needs to break into a few sections. So that’s where we really get into the RLP style tool path planning.

14:15 Where we broke it into these segments right here, which we’ll we’ll show in the print video. That first se segment on the bottom still kind of has a steeper boundary at the end of the segment. Again, just to try and keep that profile and the opening where we want it. But the sort of following segments, we can kind of control the bounding edges of that geometry to be much easier in terms of that angle tolerance that we were talking about before.

14:48 So here’s the the video of the bag printing. You can see that this is the first tool path and then we’ll see the next being one portion of the handle, the next portion, and then that final piece on top that connects it. And then we developed our silicone materials so that we could match the colors that Capernney was looking for, but still have a bag that has an additive where we can actually laser etch the Caperni branding onto each each version of the bag.

15:19 And then we were excited. We had our runway moment where we had the three versions of the bag on the runway at Disneyland Paris. And that came in three colors. The first being the aerial swipe bag with Disney and Copernney, the black version which was just Copern RLP, and finally a gray version that was just Capernium and RLP as well. And you can see it’s rigid enough to hold the shape still while we’re holding it with the things inside of it.

15:54 And of course, it’s still a flexible silicone swipe bag. And then the final piece of this sort of this collaboration with Capernie is we had talked about doing a production run of the liquid printed bag throughout the collaboration process before the show. And after the show there was indeed enough momentum that we decided to go through with it. And we printed 170 bags in all three of those colors on our own machines and that was done by our team in Boston and we shipped those off for distribution.

So that part was also rewarding for us as a team where we got the chance to do an enduse piece and in terms of the sort of more experimental versions of the swipe bags that Capernia always shows this was one of the first ones that was actually available to everyone. So that was pretty exciting for us. So I’ll go back to Hamilton now to talk about some of the software changes that came out of this.

Yeah. So through that project and all the additional projects that we have done in the past we have kind of been developing a method of understanding shape that is really applicable to our specific process. So for instance Kimble covered the angle constraint but there are a variety of other constraints that go into making something that’s manufacturable with ROP. All that specific work around the handle, around the bottom of the bag, and things like that are can be defined in a variety of different parameters.

17:21 So, you know, how does the tool path how does the wall thickness affect the radius of what’s going to be extruded? How much can you extrude on one side of the geometry versus another? How much time do you need to change between two materials in a multimaterial part? How how thin can you go? All these process parameters like that. So through all of that knowledge we’ve been able to basically craft an engine that we’ve learned about how to develop a generalized method for non-planer conformal topology first slicing which we call the RLP engine and it’s a system that allows a a user to guide their tool path generation through our system and abiding by our manufacturing constraints to achieve their design vision.

18:01 We now use this engine to achieve all sorts of parts in house through a variety of different distribution methods. So for instance for the copernney swipe bag we always start with a multi-step nonlinear analysis and optimization scheme taking in like intrinsic analysis and field optimization discovering different elements of the geometry that are relevant to our particular process. Developing constraints and setting up a initial guess also doing a feasible parameterization of the shape and a global optimization all in time before the user even gets anything back.

18:40 So then the user has the ability to go in and actually modify that. So there’s always an interaction step that allows us to change some particular things about how we want the the process to be printed. Do we want to add an additional source somewhere in the middle of the geometry? Do we want to make a circulation curve somewhere that allows the user to make it flatter in a particular area so that it abides by our constraints better than our automatic guess?

19:03 And it’s through this process where we are constraining the user to co-pilot the tool path generation and make something within our design space. There’s also a bunch of different visualizations that the user gets back in time. How does the process automatically adapt the speed of the print in order to make sure that it’s doing something of optimal quality? How can we previsualize what the tool path is going to look like before we have to calculate it?

19:29 And we also give back, okay, what’s the expected print angle and print quality of the part from an initial guess versus an optimized guess. So all of these things are going into a tool that the user has in-house and for our customers to make parts such as the copy bag. So here’s that tool path visualized. So we’ve basically chosen to develop an engine that is you know outside of any particular interface or delivery platform and we are planning on and we have delivered this in two separate manners.

20:10 One is a cloud-based manner called ver on which is for ver online and one is a grasshopper rhino manner which is for ver off ver off called ver offline. The Grasshopper method is for more advanced users like people in this room who want access to good on time. Okay. More access to you know the advanced data maybe to do something more interesting and specific for their process.

20:35 So for us you know it’s a it’s a oops it’s a double system. And then sorry this is our Grasshopper workflow. For instance, you’re inputting the mesh. You know, you’re doing a regular Grasshopper workflow. You’re visualizing different things about the mesh. What’s the expected thickness? How’s the speed control map? What’s the the tool path angles going to be? What’s the expected contours? And so on. And this is a pretty typical Grasshopper workflow that you’re going to see.

But I think what might be more important is how we deliver this to external users. People who aren’t comfortable with Grasshopper, who want to just have a button-click method of going through the process. They don’t want to have access to all those additional parameters. They just want to be on the factory floor and manufacture something that’s repeatable. They want to make 100 liners a month. They want to make a variety of different shapes.

21:30 So for instance, here we have that same editing capability, but all on the cloud. Giving back our process specific parameters as I mentioned and just clicking through the process like it’s a linear system. So we use that to you know make a variety of different parts. So vascular model here on the left which is really highlighting our gravity-free approach which is like printing completely supportless. This is the entire tool path.

21:54 There is no additional supports that are added to this. Printing directly in the air and another fun video for you guys because it’s cool. So, and printing hand models like this. So, the last piece that I’m going to emphasize is that we’ve talked a lot about software and our design services, but really truly RLP is a hardware first company. Our mission is to bring gravity-free manufacturing to the factory floor through a complete platform that includes hardware, software, and materials.

22:23 So, it’s not just a software company. We also have a Levity machine that you can buy as well. That’s like a half halfmeter cubed build volume that we announced at Rapid TCT. And it’s through all three of these things, our industrial-grade silicone with a variety of properties that you can print multiaterial with a software and a hardware that you have a new manufacturing endto-end process to deliver silicone.

And this is our team. This is who we do it with. It wouldn’t be possible without these people. So, we launched in April, our official product. We’re taking pre-orders and, we’ll be shipping our first machines in 2026. Thank you. To learn more about the CDFAM computational design symposium series, to see the archives of previous presentations, and to learn about future events, visit CDFAM.com.