top of page

Measuring Sensor Performance

Base Performance

Basic Validation Measurements for Sensor Models

Sensor models need to be compared to real world sensor data for validation. Therefore, there are high requirements on the accuracy of reference measurements to generate ground truth data needed for accurate re-simulation of the measurements.

Lidar performance measurements

We conduct real-world measurements standardized by DIN SAE Spec 91471 and compare them with simulation results to validate sensor models. Currently, we are developing semi-automated test benches for conducting these measurements.

Standardized Performance Measurements

Our test benches automatically generate ASAM OSI ground truth data, so measurements can directly be re-simulated with sensor models. The re-simulation can then be compared to the measurements with metrics provided by Persival Simspector.

Automatic Ground Truth Generation

Generated ground truth data

Weather Performance

Influence of Environment Conditions

Perception sensors are exposed to all kinds of adverse environmental conditions like rain, snow, or fog. Especially for optical sensors, e.g. lidar, the influences on the signal propagation lead to range degradation and false positive detections. These influences need to be measured and validated.

Weather test station

​Our weather test bench is equipped with precision reference sensors measuring environment parameters like visibility, brightness, precipitation drop size and speed, temperature, wind speed and direction, etc. We measure year-round in various real conditions. Multiple lidar and radar sensors can be mounted on the test bench. 

Real World Weather Measurements

In the sensors’ field of view, various stationary targets with different calibrated reflectivity are located. We analyze the weather effects in the sensor data and consider these effects in simulation. 

Calibrated Reference Targets

Calibrated reflection targets
bottom of page