Metalenz acquires PxE Holographic Imaging, expanding perception platform for consumer and Physical AI. Read the news
Metalenz

Markets · Physical AI & Robotics

New perception for
Physical AI.

Robots and autonomous machines need to know what things are made of, how they’re shaped and where they are, in any light. Metalenz delivers color, depth, and polarization with a silicon layer and semiconductor economics.

Today’s machine vision is a stack of compromises.

Machine vision systems see the world through accumulated parts: an RGB camera, a separate depth sensor, extra illumination and software fusion to stitch it together. Every added part means more cost, more calibration and more latency, and the data is still the same brightness and color every competitor trains on.

Complex Systems

Every added camera, emitter and fusion step adds cost, power and calibration.

Blind to material

Conventional cameras can’t tell glass from plastic, or skin from silicone.

Misaligned signals

Separate sensors capture the scene from different points, so data must be fused after the fact.

New Optical Intelligence with Sensor Integrated Hardware.

No fusion. No latency. At a price point machine vision has never seen.

A woman holds a highly realistic printed mask of her own face beside her face, as seen by a standard camera The same scene in polarization imaging: her real skin reads blue-green while the printed mask reads red

Polarization

Material classification and surface details.

Explore the PolarEyes platform →

RGB and Depth

Natively aligned visible and depth from a single silicon layer.

Explore the PxE platform →

Train on data no competitor can generate.

Every AI model today learns from the same billions of standard images. Metalenz hardware generates new, physics-based datasets that don’t exist anywhere else, and can only be created with the physical sensor.

The moat

No software substitute.

  • Unique information, encoded in hardware
  • Wafer-level integration at semiconductor economics
  • Proven at consumer volume: 300M+ metasurfaces deployed

Every AI model today

Billions of standard images

Cameras, phones, the internet

Brightness + color only

AI / Foundation model

Trained on the same data as every competitor

Hard ceiling: same data, same models.

vs.

Metalenz

PolarEyes

Active NIR · monochrome

PxE

Passive visible · color

Polarization Material classification Depth Phase & low light

AI with new optical senses

Trained on data no competitor can generate

New optical intelligence, absent from every existing dataset.

The moat: you cannot generate this data without the physical sensor.

Metalenz is the only company delivering across NIR and visible at semiconductor price and scale.

Simplify the system

One camera that does it all.

Today · stereo depth camera

4 discrete optical components

Diagram: stereo depth camera bar with four optical components marked 1 to 4
  • 1 RGB camera color and texture
  • 2 Left IR imager stereo pair
  • 3 Right IR imager stereo pair, set apart on a wide baseline
  • 4 IR dot projector active illumination for depth

With Metalenz

One camera: RGB, IR, depth, polarization

Diagram: compact sensor head with a single Metalenz camera marked A

Single Metalenz camera

RGB · IR · Depth · Polarization

Replaces 1–4. No stereo baseline, no projector.

A smaller, simpler sensor head for arms, grippers and mobile robots.

Applications

Equipping Machine Vision with New Understanding.

Illustration: a robotic gripper picking up a box

Piece picking & sorting

Identify and grasp transparent, reflective and mixed-material items reliably.

Illustration: a robot following a dotted path around an obstacle

Navigation

Detect glass, wet floors and low-contrast obstacles that fool standard cameras.

Illustration: a magnifying glass over a grid highlighting a surface defect

Inspection

Spot surface defects and material differences at production speed.

Let’s talk about your perception stack.

Specifications, evaluation details and roadmaps are shared under NDA. Tell us what you’re building and the right person on our team will be in touch.