Latest / The Robotics Podcast with Fexingo: Autonomous Systems, Industrial Robots, and Hardware / How Soft Robotics Changed the Factory Floor
Transcript
- Lucas: So you've probably seen those viral videos of a robot arm picking up an egg without cracking it. That's not just a party trick—it's part of a quiet shift in how robots grab things. Luna: Right, because for decades the default was basically a metal claw. Hard, fast, repeatable. But it crushes anything that isn't uniform or fragile. Lucas: Exactly. Traditional grippers—two fingers, three fingers, vacuum cups—they're great for identical metal parts on an assembly line. But the moment you hand them a tomato, or a glass bottle that's slightly different each time, you've got a problem. Lucas: Enter soft robotics. The idea is to replace rigid joints and claws with compliant materials—silicone, rubber, fabric—that conform to the object's shape without applying pinpoint force. Luna: And there's a company called Soft Robotics Inc. that's been pushing this pretty hard. They make these pneumatically actuated grippers that look like a bunch of rubber fingers that inflate and deflate. Lucas: Yeah, they call them 'mGrip' modules. Each finger is essentially a bladder. When you pump air in, the finger curls like a human finger wrapping around an object. Release the pressure and it relaxes. Lucas: The beauty is you can control the grip strength just by adjusting air pressure. So a single gripper can handle a lightbulb and then a cast-iron skillet without any tool change. Luna: But there's a trade-off, right? Speed. Pneumatic actuation is inherently slower than electric servo motors. You're not going to use these in a high-speed pick and place line doing sixty cycles a minute. Lucas: That's the big constraint. Soft grippers are generally slower—maybe ten to twenty cycles per minute. But for applications where damage prevention is worth more than speed, they win. Lucas: Take the food industry. A 2025 case study from a German bakery chain—they were using vacuum grippers to move baguettes. But the baguettes have irregular shapes, and the vacuum often lost suction or left bruises. Lucas: They switched to a soft gripper from a company called Piab—they make these 'gecko-inspired' adhesive grippers. And they cut waste by forty percent. That's huge for margins. Luna: Gecko-inspired—so not suction or clamping, but dry adhesion using micro-fibers that stick like lizard feet. That's a whole different branch of soft robotics. Lucas: Right. There's a Stanford spin-off called 'Robots With Grip' that makes those. The pads have thousands of tiny silicone stalks that create van der Waals forces. No suction, no sticky residue. Lucas: They're already used in warehouses to pick up boxes with irregular surfaces—like polybags or shrink-wrapped items that traditional suction cups can't reliably grab. Luna: Amazon piloted something similar in 2024 for depalletizing mixed items—you know, those truckloads of random products. They used a combination of soft suction and gel pads. Lucas: And the results were promising. According to a logistics trade report, they increased successful picks from around 85% to 97% on mixed sku pallets. That's a huge improvement for a system that has to handle everything from detergent bottles to bags of coffee. Luna: But let's be real—soft grippers aren't taking over the whole factory. If you're assembling a car door, you need millimeter precision and repeatable force. You're not going to use a squishy gripper for that. Lucas: No, and that's the point. Soft robotics is really about complementing traditional automation, not replacing it. The niche is any task where the object is variable, delicate, or unknown. Lucas: And there's a third category beyond pneumatic and adhesive—there's also 'jamming grippers'. Those are bags filled with granular material like coffee grounds. You press the bag onto an object, then vacuum out the air, and the granules lock together, conforming to the object's shape. Luna: I've seen those. They're surprisingly effective for irregular objects like a piece of fruit or a tool handle. But they're even slower than pneumatic fingers because you have to evacuate the air. Lucas: Yeah, the trade-off is versatility vs speed. But researchers are working on making them faster. There's a lab at Cornell that's developed a 'fast-jamming' gripper that cycles in under half a second. Lucas: If they succeed, that opens up applications in high-mix, low-volume manufacturing—which is exactly the kind of production that's growing as supply chains demand more customization. Luna: And then there's the sensing angle. A soft gripper can embed sensors in the material itself—strain gauges, pressure sensors—giving the robot a literal sense of touch. Lucas: That's the frontier. Right now most soft grippers are 'blind'—they just apply a programmed pressure. But if you add tactile feedback, the robot can adjust in real time. If it feels something start to slip, it can tighten its grip. Lucas: A company called 'RightHand Robotics' is already doing that with their piece-picking system. They use a combination of soft suction and a camera array, and the gripper has pressure sensors that tell the robot exactly how hard it's holding. Luna: So the future is basically a soft hand with a sense of touch. That's pretty wild. But how far are we from that being cost-effective for small businesses? Lucas: The cost is coming down. A pneumatic soft gripper module might run you $2,000 to $5,000, compared to maybe $500 for a simple parallel-jaw gripper. But if it saves you 40% on waste, the payback can be months. Lucas: And there's a whole ecosystem of startups now—Soft Robotics Inc., Empire Robotics, and 'Festo' has a line of soft grippers too. The competition is driving prices down. Luna: So if I'm a small manufacturer thinking about automation, where should I start? Food handling? Warehouse packing? Lucas: I'd say look at any process where you're currently using human hands because the items are too variable for a hard gripper. That's often in final packing, kitting, or inspection. Lucas: A good entry point is a collaborative robot arm—like a Universal Robots arm—paired with a soft gripper. The whole setup might cost $30,000, and it can handle tasks that previously required a full-time person. Luna: But you need to engineer the workcell carefully. Soft grippers are gentler, but they're not magic. You still need good lighting, vision, and part presentation. Lucas: Absolutely. The gripper is just one piece. The intelligence—the vision system, the planning software—that's where a lot of the value lies. But the soft gripper is what makes the physical interaction safe and adaptable. Lucas: So to wrap it up: soft robotics is solving a very specific problem—how to grab things that are fragile, irregular, or both. It's not for every factory, but in the right application, it's transformative. Luna: And it's interesting that the inspiration often comes from biology—pneumatic fingers mimic human muscles, gecko pads mimic lizard feet, jamming grippers mimic the way an elephant's trunk can grasp a peanut. Lucas: Yeah, biomimicry is a huge driver. And the next big leap will be integrating those tactile sensors so the robot can 'feel' what it's holding. That's when soft robotics really comes into its own. Luna: Alright, I'm going to go try to grip my coffee mug with a balloon. Thanks, Lucas. Lucas: Just don't inflate it too much. Thanks, Luna. That's it for this episode of The Robotics Podcast.