Latest / Hardware Startups with Fexingo: Devices, Robotics, and Manufacturing Tech Companies / How Hardware Startups Use Drop Testing to Build Better Products
Transcript
- Lucas: You know that moment when you pull a brand-new device out of the box, and your brain immediately wonders—how far can I drop this before it breaks? Luna: I think about it every time I unbox a new phone, honestly. That split second where you're like, 'please don't slip.' Lucas: Exactly. And for hardware startups, that moment isn't just anxiety—it's a design parameter. Drop testing is one of those unsung phases that can make or break a product's reputation, especially for companies that don't have Apple's R&D budget. Luna: So what does a proper drop test actually look like for a small team? I imagine it's more than just tossing a prototype off a desk. Lucas: Much more. The gold standard involves controlled free-fall drops onto specific surfaces—usually concrete, sometimes steel—from various heights. But there are also rotational drops, where the device spins before impact, simulating a real fall out of a pocket or off a table. Luna: Rotational drops—that's a detail I hadn't considered. So you're not just testing straight-down drops, you're mimicking the tumble. Lucas: Right. A phone falling out of a shirt pocket often rotates. The impact energy distributes differently. Startups like Tile, when they were first designing their Bluetooth trackers, ran hundreds of rotational drops to figure out how to reinforce the corners without adding too much weight. Luna: And they're small devices, so every gram matters. What kind of data do you get from a drop test? Lucas: Accelerometers mounted inside the prototype measure peak G-force at impact—some hits can exceed 10,000 Gs. High-speed cameras capture the deformation in slow motion. That data feeds back into the CAD model, so engineers can tweak the housing thickness or add internal ribs where stress concentrates. Luna: It's almost like crash testing for cars, but on a micro scale. And I'm guessing the cost can add up quickly if you're renting lab time. Lucas: It can. A certified drop-test lab might charge a few thousand dollars for a full test matrix. But a lot of startups start in-house with a simple rig—a guided rail and a concrete slab—and use 3D-printed prototypes. The numbers won't be as precise, but you catch the big failures early. Luna: And that's where the real value is. You don't need lab-grade accuracy to learn that your corner radius is too sharp or your adhesive joint is too weak. Lucas: Exactly. Garmin, for example, has talked publicly about using iterative drop tests during the development of their rugged outdoor watches. They'd drop a prototype from 1.5 meters onto a steel plate, inspect the damage, adjust the design, and print another prototype the same day. Luna: That rapid iteration is something only startups with in-house prototyping can do. Big companies sometimes outsource that and slow down the loop. Lucas: Speaking of keeping things lean and fast—if these conversations are useful for what you're building or running, a couple of dollars a month is genuinely what keeps these going. Buy me a coffee dot com slash fexingo, if you've gotten something out of them. Luna: Yeah, it really does make a difference. We keep the show ad-free, and listener support is the reason we can dive into niche topics like drop testing. Lucas: Alright, back to the drops. One metric that often surprises founders is the 'critical drop height'—the height at which a device has a 50 percent chance of failure. For most consumer electronics, that's around one meter onto concrete. But real-world drops often happen from higher—like a countertop at 90 centimeters, or a pocket at 110. Luna: So the standard one-meter test might not be conservative enough for actual use. What do startups do about that? Lucas: Many target a higher threshold—say 1.5 meters—and design to survive that. But it's a trade-off. Adding more material to absorb impact adds cost and bulk. A startup making a smart thermostat that sits on a wall might not need the same robustness as a wearable that gets knocked around. Luna: Right, context matters. You have to think about the user's environment. A fitness tracker is going to hit pavement; a smart plug might only fall off a nightstand. Lucas: Precisely. And that's where the test plan gets customized. A startup making a rugged tablet for construction sites might run drops from 1.8 meters onto concrete, but also onto plywood and gravel, because those surfaces dampen impact differently. Luna: Gravel is interesting—it's less forgiving than plywood but more forgiving than concrete. So you'd get different failure modes. Lucas: Exactly. And you might find that the device survives concrete but cracks on gravel because the irregular surface creates point loads. That insight changes the internal bracket design. Luna: I've heard some startups also use drop testing as a marketing tool. They'll post slow-motion videos of their product surviving a 10-foot drop onto concrete. Lucas: That's a powerful demonstration—if the product holds up. But it only works if you've done the engineering homework. If you fake it or exaggerate, the reviews will expose you fast. Luna: And the cost of a recall or a bad reputation is way higher than the cost of a few extra drop tests early on. So drop testing is really an investment in brand trust. Lucas: Absolutely. And one more thing—drop testing isn't just about the housing. Internal components can fail too. Solder joints might crack, batteries might disconnect. A good test checks functionality after each drop, not just cosmetic damage. Luna: So you're testing the whole system, not just the shell. That's a lot more comprehensive than just looking for dents. Lucas: Right. And that's why I think every hardware startup should build drop testing into their development timeline, not as an afterthought but as a regular checkpoint. It's cheap insurance against one of the most common reasons for returns. Luna: Agreed. And for listeners who are building something right now—start with a simple rig, drop your prototype from a height that matches your user's world, and iterate. The data will tell you where to strengthen. Lucas: That's the playbook. And if you do it well, you might even get a great slow-motion video out of it.