# How Chip-Scale Optical Phased Arrays Steer Light Without Moving Parts

An episode of [The Hardware Podcast with Fexingo: Chips, Devices, and Electronics Engineering Conversations](https://invisiblenetwork.ai/podcast/1896799982), 4 Aug 2026, 9:24.

Audio: https://audio.fexingo.com/business/the-hardware-podcast/episode-0145.mp3

Lucas and Luna dive into optical phased arrays, the chip-scale technology that steers laser light electronically with no moving parts. They trace how this tech, once confined to military radar, is now enabling solid-state LiDAR for autonomous vehicles and free-space optical communications. The episode explains the physics - how tiny phase shifters create constructive interference to direct a beam - and why it beats spinning mirrors. They cover the manufacturing challenges, the role of silicon photonics, and how costs are falling as foundries scale. You'll learn about the companies racing to commercialize it, the bet on solid-state LiDAR for self-driving cars, and the latest breakthroughs from summer 2026. If you've ever wondered how your car will see without moving parts, this episode delivers the specifics. Brought to you by Fexingo Business Podcast Network. #OpticalPhasedArrays #SiliconPhotonics #LiDAR #AutonomousVehicles #Photonics #SolidStateLiDAR #BeamSteering #FreeSpaceOptics #ChipScale #PhasedArrays #LaserTechnology #Semiconductors #Technology #Engineering #Innovation #FexingoBusiness #BusinessPodcast #Hardware Keep every episode free: buymeacoffee.com/fexingo

## Transcript

Lucas: So there's a technology that's been around for decades in military radar, but it's only now, in 2026, really hitting its stride in consumer and automotive applications. I'm talking about optical phased arrays.
Luna: Optical phased arrays. That's the thing that steers a laser beam without any moving parts, right? I've heard the term but I'm fuzzy on the details.
Lucas: Exactly. No spinning mirrors, no gimbals. Just a chip with a bunch of tiny antennas that delay light in precise ways to point the beam wherever you want. And it's becoming the backbone of solid-state LiDAR for self-driving cars and even satellite to satellite communication.
Luna: Before we go deeper, I have to say, if you're enjoying these conversations as much as I am, it's worth considering a small coffee for the show. It's what keeps this podcast ad-free and independent. The link is buy me a coffee dot com slash fexingo.
Lucas: And even a one-time coffee genuinely helps us keep digging into these topics. But back to the optical phased array - let's break down what it actually is.
Lucas: The core idea is that if you have an array of tiny optical antennas, each emitting the same light, you can adjust the phase of each one. When the phases align in a certain direction, the light waves constructively interfere and create a strong beam pointing that way. Change the phases, and the beam moves.
Luna: So it's like a radio phased array, but for light. And the wavelengths are so much shorter, so the antennas have to be microscopic, right?
Lucas: Yes. For near-infrared, say 1550 nanometers, which is what these systems often use, the emitter spacing has to be about half that wavelength to avoid grating lobes. That's around 775 nanometers. So you're packing thousands of these emitters onto a chip.
Luna: And the phase shifters - they're the heart of it. How do they work in silicon photonics?
Lucas: The most common approach is thermo-optic. You heat a tiny section of the waveguide with a micro-heater, and because silicon's refractive index changes with temperature, the light slows down a bit. That's the phase delay. It's simple and reliable, but it consumes power and it's a bit slow.
Lucas: The faster alternative is electro-optic, using things like lithium niobate, which changes its refractive index in response to an electric field. That's much faster, but integrating lithium niobate with standard silicon manufacturing is still tricky.
Luna: I've read that the real challenge is calibration. With thousands of phase shifters, you can't just set them once. Thermal drift and variations in the chip mean you need constant feedback.
Lucas: That's exactly the hard problem. Each emitter has a slightly different phase error, and the errors drift as the chip warms up. So you need a calibration algorithm that periodically tweaks the phase settings to keep the beam sharp. Some systems use an integrated photodiode to monitor the output and adjust in real time.
Luna: And this is what's been holding back commercial adoption, right? The complexity.
Lucas: For sure. But in the last couple of years, advances in packaging and control electronics have made it much more practical. We're seeing companies like Quanergy and others pushing solid-state LiDAR based on optical phased arrays, aiming for the automotive market.
Luna: I've also seen a bunch of startups in the free-space optical communication space, like for last-mile internet and inter-satellite links. They love the speed and the lack of moving parts.
Lucas: Right. When you're pointing a laser from one satellite to another moving at thousands of kilometers per hour, you can't afford a motor that takes seconds to slew. An optical phased array can steer in microseconds. That's a game-changer.
Luna: And what about the power consumption? I imagine heating thousands of phase shifters gets hot.
Lucas: It does. In a typical thermo-optic array, each heater might draw tens of milliwatts. If you have a thousand elements, that's potentially tens of watts, which is a lot for a battery-powered device. That's why there's a push toward more efficient designs, like using a sparse array or only turning on a subset of elements for a given direction.
Lucas: There's also the beam quality. With a small aperture, the beam is wide, which limits the resolution. To get a narrow beam like a mechanical system, you need a large array, but then you run into yield issues and cost.
Luna: So what's the sweet spot for today's chips? How many elements are we talking?
Lucas: Some research prototypes have shown arrays with 512 elements, and a few companies have demonstrated chips with over a thousand. But for production, it's often fewer, like 128 to 256, and they use a technique called optical phased array plus a lens to get a decent field of view.
Luna: And the field of view - that's another limitation, right? You can't steer arbitrarily wide.
Lucas: Typically, you can steer over maybe 60 degrees in one dimension, and the other dimension might be handled by a different mechanism, like a grating that emits light at different angles depending on wavelength. So you can do wavelength tuning to steer in one direction and phase tuning in the other.
Lucas: That's called wavelength-selective steering, and it's a clever way to get 2D scanning with just a 1D array. But it means you need a tunable laser, which adds cost.
Luna: I've also heard about using optical phased arrays for 3D sensing in AR glasses, like tiny projectors that paint a pattern on the world.
Lucas: Yes, that's a promising area. Instead of spinning mirrors, you can scan a structured light pattern with a chip the size of a thumbnail. That could enable compact depth sensors for AR, and it's being explored by quite a few labs.
Luna: So where does the industry stand in mid-2026? Have we seen real deployments or is it still mostly research?
Lucas: I'd say we're right at the cusp. There are several companies claiming to have solid-state LiDAR units based on optical phased arrays that are sampling to automotive customers. The key is whether they can hit the cost targets - under a hundred dollars per unit for high volume.
Lucas: And I've seen a few announcements this summer about manufacturing partnerships with silicon photonics foundries, which is a sign that they're moving from prototypes to production.
Luna: It's fascinating because it's such a departure from the mechanical LiDAR that dominated the early autonomous vehicle tests. Those spinning units cost tens of thousands of dollars.
Lucas: Exactly. The promise is that an optical phased array can be mass-produced like any other silicon chip, so the cost drops dramatically. It's a classic semiconductor play - once you've got the design right, the marginal cost of another unit is just the silicon.
Luna: I've also read about challenges with environmental robustness. Like, how does it handle shock and vibration, temperature swings?
Lucas: That's actually one of the advantages. With no moving parts, it's inherently more robust and can handle vibration well. The main concern is thermal management because the phase shifters generate heat, and the calibration has to adapt to temperature changes.
Lucas: But advances in on-chip temperature sensors and real-time control loops are addressing that. Some designs even use a reference interferometer to measure the temperature and adjust the phase settings accordingly.
Luna: So it's not just a photonics problem, it's a control systems problem.
Lucas: Absolutely. The chip is half the story, the firmware is the other half. And that's where a lot of the IP is being built up.
Lucas: One more thing - the free-space optical communication angle. With the rise of low-earth-orbit satellite constellations, there's a real need for high-bandwidth laser links between satellites. And optical phased arrays are a natural fit because they can steer quickly and precisely without bulky gimbals.
Luna: I've seen some companies like Mynaric and others working on this for years. Is it finally becoming practical?
Lucas: Yes, I think we're seeing the first operational links this year. The latency and bandwidth benefits are compelling. And because these satellites are in LEO, they move fast relative to each other, so you need that fast steering.
Luna: With all these applications, it feels like optical phased arrays are one of those technologies that's been 'five years away' for a while. But maybe this time it's actually here.
Lucas: I think so. The combination of mature silicon photonics, better calibration algorithms, and clear market pull from autonomous vehicles and satellite communication has pushed it over the edge. In another couple of years, I wouldn't be surprised if every self-driving car has one of these chips inside.
Luna: So the next time we see a car without a spinning top, we'll know it's probably using an optical phased array.
Lucas: Exactly. It's one of those quiet revolutions, where the tech disappears into the product. And that's a good thing.

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.