An engineer named Gauthier Lechevalier has built a working prototype of a radar system that can tell what objects are made of — something ordinary radar cannot do.

Normal radar works like this: it sends out radio waves, they bounce off an object, and from the echo the radar can figure out where the object is and how fast it's moving. Lechevalier's system does something harder. It analyzes the echo in fine detail to figure out what material the object is made from — whether it's metal, plastic, or something else.

He did this by combining two techniques. The first is called beamforming. Think of a radar antenna like a flashlight: ordinary radar has a flashlight that shines in all directions at once. Beamforming is like adding a lens that narrows the light into a tight beam pointed in one specific direction. This concentrated beam means the radar's echo is cleaner and easier to read. The second technique is special software that analyzes the echo in detail to identify the material.

To identify what something is made of, you have to look for subtle clues in how the radio wave bounces back — things like how easily the material lets electricity move through it, whether its surface is rough or smooth, and how much radio energy it absorbs. All of these details change the echo in ways that software can recognize.

Why does this matter? Full material identification normally requires expensive equipment. Lechevalier's approach is cheaper and can see through clothing or packaging — which makes it useful for checking product quality in factories, inspecting baggage for security, or monitoring the condition of buildings and bridges.

The reason Lechevalier could build this at all is that radar chips have become affordable. Over the past ten years, the price of millimetre-wave radar chips dropped enough that independent engineers can experiment with them. When hardware got cheaper, the focus of the engineering work shifted from building custom antennas to writing better software algorithms. That is where the real work is in Lechevalier's prototype.

Here's something important to keep in mind: a working prototype in a laboratory is different from a system that works reliably in the real world. In a factory with controlled conditions and known materials, identifying what something is made of is straightforward. But when a system has to work in messy, changing conditions — with different shapes, different materials, temperature changes, and background noise — the job becomes much harder. Lechevalier's documentation does not say how well this system performs across different materials and conditions, so those practical questions are still open.

What the prototype does show is that this combination of focused radar beams and smart analysis is possible with affordable hardware. That opens a door. In the coming years, you might see similar radar systems built into factory equipment to check product quality, into waste-sorting machines, or into sensors that monitor the health of buildings and bridges. Radar has been gaining new capabilities steadily: it started measuring just speed, then it recognized hand gestures, then it detected whether a room was occupied, and now it is learning to identify materials. Each step has followed the same path: first it's specialized and expensive, then it becomes accessible to engineers and hobbyists, and finally it gets built into real products.