Latest / Hardware Startups with Fexingo: Devices, Robotics, and Manufacturing Tech Companies / Why Hardware Startups Are Using Hybrid Additive Manufacturing
Transcript
- Lucas: So you're building a hardware prototype — something with metal parts that need to be both complex and precise. Traditionally you'd either 3D print it, which gets you geometry but not surface finish, or you'd CNC it, which gives you precision but takes forever and wastes material. Luna: And then you have to ship the part between the printer and the mill, re-fixture it, hope the alignment holds. That's easily a week lost per iteration. Lucas: Exactly. That's the pain hybrid additive manufacturing is built to solve. One machine — it prints a near-net shape, then switches to a milling spindle to machine the critical surfaces to tolerance. No transfer, no alignment error. Luna: If this kind of practical, operator-level insight is useful to you — and I know it is for a lot of our listeners who are actually building things — these episodes stay ad-free because of listener support. You can find that at buy me a coffee dot com slash fexingo. Lucas: Yeah, genuinely. That support lets us keep digging into these specific manufacturing tactics. So — back to the hybrid approach. The big idea is that you get the design freedom of additive with the tolerances of subtractive, all in one setup. Luna: And this isn't just a lab curiosity anymore. Companies like DMG MORI and Markforged have had hybrid machines on the market for a few years now. What's changed that makes it relevant for a startup? Lucas: The price point has come down — you can get a desktop hybrid that handles aluminum and mild steel for around fifty thousand dollars now. That's not cheap, but compare it to buying a separate printer and a separate mill, plus the labor to transfer parts back and forth. Luna: Right, the total cost of ownership might actually favor hybrid for small batches. What kind of timeline savings are we talking? Lucas: I talked to a robotics startup out of Pittsburgh — they make collaborative arm actuators. Their old process: print the gear housing on an SLS machine, ship it to a local machine shop, wait for them to drill the mounting holes and bearing seats. That was three weeks per iteration. Luna: Three weeks. And how many iterations do you need before you freeze the design? Five, six? Lucas: They went through eight before they were satisfied. So that's twenty-four weeks just in prototyping. After they bought a hybrid machine from one of the Japanese builders — I won't name the exact model because they've since upgraded — they got that same part done in four days. Print overnight, machine the next morning, test by afternoon. Luna: Four days. That changes your whole development cycle. You can try more designs, fail faster. Lucas: Exactly. And the material waste is dramatically lower. With pure CNC, you start with a block and cut away maybe seventy percent of it. With hybrid, you deposit only the material you need and then machine off just the last few millimeters. That's a huge cost saving if you're using something like titanium or Inconel. Luna: But the trade-off is the machine itself is more complex. More axes, more things that can break. And the software toolchain isn't totally seamless yet. Lucas: That's fair. You're essentially running two processes in one enclosure, so the CAM software has to handle both additive path planning and subtractive toolpaths. The leading platforms are getting better — Autodesk Fusion 360 now has a hybrid module, and Mastercam added support last year. But it's not plug and play. You need someone who understands both processes. Luna: So the human capital requirement is higher. That might be the real barrier for a two-person startup. Lucas: It can be. But I've seen founders who came from a machining background pick it up quickly. And if you're designing parts that are impossible to machine conventionally — internal channels, lattice structures, conformal cooling — then hybrid is almost your only option. Luna: Let's talk about a specific application where hybrid really shines. Conformal cooling channels in injection molds, for example. Luna: So if you're a hardware startup making your own injection molds — say for a consumer electronics enclosure — hybrid additive can get you to production faster and with better part quality. Lucas: Exactly. And the mold itself is one part, not multiple pieces bolted together. That reduces leak paths and maintenance. The ROI calculation becomes pretty compelling when you factor in the longer mold life and faster cycles. Luna: But what about surface finish? Even after machining, do hybrid parts have the same properties as a fully wrought billet? Lucas: That's the key question. With laser powder bed fusion, the as-printed surface is rough — maybe Ra 10 to 20 microns. After machining, you can get down to Ra 0.8 or better. The bigger concern is porosity in the printed core. If the additive layer isn't fully dense, you can get micro-voids that reduce fatigue life. Luna: So you need to qualify the process. Run tensile tests, do CT scanning on early parts. That adds time and cost. Lucas: It does. But for non-critical structural parts — brackets, housings, fixtures — the risk is low. And the standards are emerging. ASTM is working on a specification for hybrid additive-subtractive processes, and the FAA has already certified some hybrid-printed parts for aerospace interiors. Luna: Let's bring it back to the startup context. If I'm a founder with a hardware concept, when should I even consider investing in a hybrid machine versus just using a service bureau? Lucas: Good question. The rule of thumb I've heard from several founders: if you're going to need more than twenty iterations of a metal part that requires both complex geometry and tight tolerances, buy the machine. If it's fewer than that, use a service like Xometry or Protolabs — they offer hybrid as a process option now. Luna: Twenty iterations. That's a pretty specific threshold. What about the cost per part? Lucas: At a service bureau, a hybrid part in aluminum might run you two hundred to four hundred dollars, depending on size and complexity. The material cost in-house is maybe twenty dollars. So after fifty parts, you've paid for the machine. But that's just the direct cost — you also save the lead time, which for a startup racing to market is often more valuable. Luna: That makes sense. So the decision is really about velocity. Do you need to iterate fast and control your own schedule? Lucas: Exactly. And there's a secondary benefit: learning. When you own the machine, you understand the process limitations intimately. That knowledge feeds back into your design decisions. You stop designing parts that are hard to make and start designing parts that are easy to make with hybrid. Luna: That's the DFM — design for manufacturability — angle. It's not just about making parts, it's about learning how to design better parts. Lucas: Precisely. And the technology is still evolving. Some newer hybrid machines incorporate in-process monitoring — sensors that measure melt pool temperature or layer thickness, and the machine adjusts parameters in real time. That reduces the need for post-process inspection. Luna: So we're moving toward a future where the machine itself ensures quality, and the operator just loads the design file and presses start. Lucas: That's the vision. But today, hybrid additive manufacturing is a tool that rewards hands-on engagement. For a hardware startup willing to invest the time, it can collapse months of development into weeks. Luna: And with the cost coming down and the software maturing, it's becoming accessible to more than just the deep-pocketed players. Lucas: Right. So if you're a founder staring at a prototype that's taking forever to get machined, it might be worth a look. Just run the numbers on your specific part volume and iteration count. Luna: And maybe start with a service bureau to validate the process before you buy the iron. Lucas: Smart approach. And that's what we've seen successful startups do — test, validate, then commit to the capital expenditure. The ones that skip the validation step often regret it. Luna: Great advice. Thanks, Lucas. That's all for this episode of Hardware Startups with Fexingo. Lucas: Thanks, Luna. See you next time.