# How Chip-Scale Force Sensors Are Making Robots Feel

An episode of [The Hardware Podcast with Fexingo: Chips, Devices, and Electronics Engineering Conversations](https://invisiblenetwork.ai/podcast/1896799982), 14 Aug 2026, 11 min.

Audio: https://audio.fexingo.com/business/the-hardware-podcast/episode-0155.mp3

In episode 155 of The Hardware Podcast, Lucas and Luna explore the surprisingly tactile world of chip-scale force sensors—the tiny silicon devices that let robots, surgical tools, and even smartphones feel pressure, grip, and touch. They open with a specific moment: a robot arm in a lab gently picking up an egg without cracking it, made possible by a sensor that measures micro-Newtons of force. Lucas breaks down the physics—piezoresistive and capacitive designs, MEMS fabrication, and the challenge of measuring force without adding bulk or cost. Luna digs into the clinical angle: how force-sensing catheters are changing cardiac ablation procedures, and why haptic feedback in prosthetics is the next frontier. They touch on the business side too, from the incumbent players like Tekscan to the startups racing to commercialize tactile sensors for robotics in 2026. The conversation stays grounded in engineering specifics—bridge circuits, drift, calibration—while connecting to the bigger promise: machines that can feel the world as precisely as we do. A pragmatic look at a quiet revolution in sensing. #ForceSensors #MEMS #Robotics #Haptics #MedicalDevices #SurgicalTools #Prosthetics #Piezoresistive #CapacitiveSensing #TactileSensors #RobotGrippers #SensingTechnology #Hardware #Engineering #TechPodcast #FexingoBusiness #BusinessPodcast #Innovation Keep every episode free: buymeacoffee.com/fexingo

## Transcript

Lucas: So there's a video that's been doing the rounds internally at a few robotics labs: a robot arm, this big industrial-looking thing, picking up an egg. Not crushing it, not dropping it, just lifting it gently and setting it down. And the reason that's not just a party trick is that underneath the gripper there's a chip that measures force in the micro-Newton range.
Luna: Micro-Newtons? That's like the weight of a tiny dust particle, right?
Lucas: Exactly. A micro-Newton is roughly the force exerted by a single grain of salt resting on a surface. And that's the kind of sensitivity we're now getting from chip-scale force sensors. These are devices you can fit on the tip of a fingertip, yet they can resolve forces that are essentially imperceptible to a human hand.
Luna: So how do you actually build something that small that can measure force that precisely? Is it a strain gauge thing?
Lucas: Right, strain gauges are the classic approach, but the chip-scale version usually relies on one of two principles. The first is piezoresistive: you have a tiny silicon diaphragm, and when it deforms under pressure, its electrical resistance changes. The second is capacitive: you have two plates, and the distance between them changes as force is applied, which alters the capacitance. Both can be fabricated using standard MEMS processes, so you can shrink them down to a few square millimeters.
Luna: And the signal you get out of that—is it clean enough to use in real-time control? Because a robot that's trying to grip something needs feedback fast, or it's going to overshoot.
Lucas: That's the hard part. The raw signal is tiny, so you need careful amplification and filtering. But the bigger challenge is drift and calibration. A force sensor that reads zero when there's no force, but drifts over time or with temperature, is useless. That's why you see a lot of designs with built-in reference capacitors or resistors to compensate for environmental changes.
Luna: So what's actually driving the demand now? I mean, we've had force sensors for decades, but they were bulky. What's changed that makes chip-scale so interesting?
Lucas: Three things. Robotics is the big one—humanoid robots and collaborative robots need to handle objects that vary in shape and fragility. A robot that can only pick up rigid boxes is not going to fold laundry or assemble a smartphone. The second is medical devices, like catheters that can sense contact forces during cardiac ablation. And the third is consumer electronics, where haptic feedback in phones and wearables is getting more sophisticated.
Luna: Let's stick with medical for a second. I remember reading about force-sensing catheters for heart surgery. How does that work?
Lucas: So during cardiac ablation, a doctor threads a catheter through a blood vessel up to the heart, and then uses radiofrequency energy to burn a tiny spot to correct an arrhythmia. The tricky part is making sure the tip is in proper contact with the tissue. Too little contact and the burn isn't effective; too much and you risk perforating the heart wall. A chip-scale force sensor at the tip can measure contact force in real-time, giving the doctor feedback they never had before.
Luna: That's incredible. So you're essentially putting a sense of touch into a surgical instrument. And the same kind of technology is going into prosthetics, right?
Lucas: Exactly. Prosthetic hands are starting to integrate force sensors in the fingertips, so the user gets a sense of how hard they're gripping. That's a huge deal for someone who wants to hold a paper cup without crushing it, or shake hands without being too limp. And we're also seeing tactile sensors in robotic skin, which is a whole other beast—covering a whole robot body with a distributed array of force sensors.
Luna: I imagine that's expensive. How does that affect the market? Like, who's actually selling these things, and what's the price point?
Lucas: The market is still pretty niche. Established players like Tekscan have been around for decades, but their sensors are more like flexible films rather than chip-scale. The newer entrants are startups that are using MEMS foundries to produce force sensors in volume. Prices are coming down, but a single high-accuracy chip-scale force sensor can still cost anywhere from a few dollars to tens of dollars, depending on the specs. For a robot with dozens of sensors, that adds up.
Luna: So it's not just about the sensor itself; it's about the whole system—the wiring, the data processing, and the algorithms that interpret the signals. That's probably where the real value is.
Lucas: Absolutely. And that's why you see companies like Force Dimension, or even some of the big semiconductor guys, getting involved. The sensor is one component, but the real moat is in the software that turns raw capacitance changes into a meaningful grip force. That's where the engineering gets really interesting.
Luna: I mean, you've got to wonder about the calibration. How do you even test a sensor that's measuring micro-Newtons? It's not like you can just hang a weight on it.
Lucas: That's a good question. For calibration, you use very precise force sources, like electrostatic force generators or micro-machined weights that are so small they're essentially fabricated with the sensor. Some labs use a setup called a 'force probe' that applies a known force using a piezoelectric actuator. And then you have to do this over a range of temperatures and humidity levels, because those affect the mechanical properties of silicon.
Luna: So it's not just a chip design problem, it's a metrology problem. That's probably why there's no single company that's blown everyone away yet.
Lucas: Right. And there's also the question of dynamic range. A sensor that can measure a micro-Newton might saturate if you hit it with a kilogram of force. You need a wide range, and that's tough to achieve in a single chip. Some designs use a combination of a stiff and a soft structure to get a broader range, but that adds complexity.
Luna: So where do you see this going in the next few years? I mean, we're in August 2026. Are we going to see these in every robot, or is it still going to be a premium feature?
Lucas: I think we're at a tipping point. The cost curve is dropping, and the demand from both robotics and medical is strong. But the big opportunity might be in something like 'smart grippers' for warehouse automation. If you can make a sensor that's reliable, cheap, and can handle the demands of a logistics center, that's a massive market. Plus, with the rise of humanoid robots, the need for tactile feedback is becoming non-negotiable.
Luna: I've also heard about force sensors being used in sports equipment, like smart insoles that measure foot pressure. Is that a real thing?
Lucas: It's real, and it's growing. Companies like Moticon and Novel have been making pressure-sensing insoles, and now they're integrating chip-scale sensors to reduce weight and improve resolution. That opens up not just sports, but also diabetic foot care, where you need to monitor pressure points to prevent ulcers. It's another example of how a tiny sensor can have a big impact.
Luna: That's fascinating. And before we go too far, I know we've got a really supportive listener base that helps keep this show going without ads. If you're one of the people who chips in a few bucks a month, we genuinely appreciate it. It's what allows us to dive into topics like this.
Lucas: Yeah, and if you're not, no pressure. But every little bit helps, and you can do that through buy me a coffee dot com slash fexingo. It literally pays for the time we spend researching these episodes.
Luna: Yeah, it's a small thing that keeps this whole thing running. So thanks to those who do. Now, back to the tech—you were talking about the future of force sensing. What's the one development you're most excited about?
Lucas: I think it's the integration of force and proximity sensing into a single chip. Imagine a robot hand that can sense an object before it even touches it, and then measure the exact force as it makes contact. That would be a game-changer for manipulation tasks. There are already research prototypes that combine capacitive force sensing with time of flight proximity sensing on the same die.
Luna: So it's like giving a robot both the sense of touch and a primitive sense of vision on the fingertips. That's wild.
Lucas: Exactly. And it's not just for robots. The same technology could be used in augmented reality, where you want to sense how a user is gripping a controller or a wearable. Or in automotive, to detect if a driver's hands are on the wheel and how firmly. The applications are really broad.
Luna: So it's a classic situation where a sensor that used to be a niche instrument is becoming a commodity component. And the winners will be the ones who can integrate it well with the rest of the system.
Lucas: Right. The sensor is the easy part in some ways. The hard part is making it robust, reliable, and manufacturable at scale. And that's where the real engineering challenge lies. So I'd say keep an eye on the companies that are shipping actual products, not just lab demos. The ones that can do that are going to define the next decade of robotics and medical devices.
Luna: And for our listeners who are engineers or just curious, where can they learn more? Are there any open-source projects or standards emerging?
Lucas: There's actually a growing open-source ecosystem around tactile sensing. The Tactile Sensor Library is a collection of designs and firmware that you can use to build your own force sensors. And there's the IEEE Tactile Sensing Committee, which is pushing for standardization. If you're interested, that's a great starting point.
Luna: That's awesome. So it's not just a closed world of proprietary tech. There's a community building around it.
Lucas: Definitely. And that's how we get from the egg-picking demo to mass adoption. It's the same pattern we've seen with MEMS microphones and accelerometers—they started as high-end components and then became ubiquitous. Force sensors are on that same trajectory.
Luna: I love that. So next time I see a robot picking up an egg, I'll know there's a tiny chip doing the heavy lifting.
Lucas: Exactly. And it's not just eggs. It's surgical tools, prosthetics, even the keyboard on your laptop. Force sensing is quietly becoming a part of everything we touch.

Shows are listed from Apple Podcasts' charts and directory, and Podcast Index where it is used. Every show and episode belongs to its publisher and plays from the publisher's own feed; this site hosts none of them.