Building Realism: Our Sensor Material Database for Lidar Simulation

Simulating lidar sounds straightforward at first: send out light, measure the distance, and construct the point cloud. In practice, however, the challenge lies in what happens in between. The way infrared light interacts with the world is anything but simple, and without modeling that interaction correctly, even the best sensor model will fall short.

Simulated point cloud with different materials

At the core of this problem is reflectivity. Every material reflects light, especially infrared light, differently, and that behavior depends on several factors: the wavelength of the lidar, the viewing geometry, and, of course, the material itself. What might seem like a simple surface can behave very differently depending on how you observe it.

This is exactly why we built our sensor material database.

Explore the Database

Measuring the real world

Rather than relying on approximations or static reflectivity values, we measure how materials actually behave. Our patented measurement device is designed specifically for lidar applications. It uses two rotating arms, one carrying the laser source, the other the detector, allowing us to capture the interaction between light and material from many different perspectives.

By supporting wavelengths from 905 nm up to 1550 nm, the system covers the range used by modern lidar sensors. More importantly, it lets us observe how reflectance changes not just with wavelength, but also with the angle of incidence and observation. The result is a full, wavelength-dependent bidirectional reflectance distribution function (BRDF) for each material.

Instead of a single number, you get a complete description of how a surface reflects light.

Persival Reflectivity Measurement Device

Beyond “diffuse” and “specular”

In many simulation setups, materials are simplified into categories like diffuse, specular, or retroreflective. While this is convenient, it rarely reflects reality.

Most materials don’t belong to a single category. A surface that looks purely diffuse at one angle can suddenly show strong specular characteristics at shallow incidence. Likewise, materials engineered for visibility, like lane markings or traffic signs, often combine multiple reflection mechanisms that depend heavily on viewing conditions.

Capturing these transitions is crucial for realistic simulation. It directly impacts how objects appear in point clouds, how strong returns are, and how reliable detection becomes in edge cases.

Example BRDF measurement of a cobblestone

A database grounded in real materials

Over time, we have built up a growing database of measured materials that are directly relevant for automotive and robotics applications. It includes common road surfaces such as different kinds of asphalt, along with concrete and pavements like cobblestone. Various lane markings, both temporary and permanent, are covered, as well as traffic signs with their often highly engineered reflective properties.

Beyond infrastructure, the database also includes a wide range of automotive paints, fabrics in different colors and textures, reflective vests, as well as materials of NCAP dummy targets. The goal is not just breadth, but realism: capturing the diversity of materials that sensors encounter in the real world.

You can explore the growing database here.

The data is not just descriptive, but immediately usable: the measured BRDFs can be directly integrated into simulation tools such as dSPACE AURELION and IPG CarMaker, or exchanged via the ASAM OpenMATERIAL 3D standard. This ensures that the same physically accurate material behavior can be used consistently across different simulation environments.

Tailored to your environment

Standard materials are only part of the picture. In many cases, what really matters is the exact surface in your specific environment. The asphalt on a proving ground, the coating used on a facility, or custom materials in a particular deployment can all influence sensor behavior.

That’s why we also offer custom measurements. If you need accurate simulation for a specific setup, we can measure those exact materials and integrate them into the database. This makes it possible to move from generic assumptions to environment-specific accuracy.

Bridging simulation and reality

High-quality lidar simulation depends on high-quality input data. By combining controlled measurements, multi-wavelength coverage, and full angular resolution, our sensor material database provides a solid physical foundation for simulation.

It helps close one of the biggest gaps between simulated and real-world perception: the behavior of infrared light itself.

Because in the end, realistic simulation is not just about modeling sensors, it’s about modeling the world they see.