# A Benchmark for Lidar Sensors in Fog: Is Detection Breaking Down?

**Authors:** Mario Bijelic, Tobias Gruber, Werner Ritter  
**arXiv:** [1912.03251](https://arxiv.org/abs/1912.03251)  
**Date:** 2019-12-06  
**Summary page:** https://www.summarizepaper.com/en/paper/1912.03251v1/

## Key points

- Achieving level five autonomy in autonomous driving requires more than just self-driving capabilities under ideal weather conditions.
- Adverse weather such as fog, rain, and snow degrade the perception of the environment for autonomous vehicles.
- A study was conducted to evaluate the performance of current lidar sensors in foggy conditions.
- Four different lidar systems were tested in a controlled environment within a fog chamber.
- The researchers aimed to identify problems and disturbance patterns associated with these lidar systems in adverse weather conditions.
- Tuning internal parameters could potentially enhance the performance of these sensors in bad weather situations.
- Understanding limitations and disturbances when using lidar sensors in adverse weather helps develop more robust algorithms for accurate perception.
- Evaluating lidar sensor performance in adverse weather scenarios is important for achieving higher levels of safety and reliability.

## Summary

In the field of autonomous driving, achieving level five autonomy requires more than just self-driving capabilities under ideal weather conditions. Adverse weather such as fog, rain and snow pose significant challenges as they degrade the perception of the environment for autonomous vehicles. To address this issue, a study was conducted to evaluate the performance of current state-of-the-art light detection and ranging (lidar) sensors in foggy conditions. The study involved testing four different lidar systems in a controlled environment within a fog chamber. The researchers aimed to identify the current problems and disturbance patterns associated with these lidar systems when operating in adverse weather conditions. Additionally, they explored how tuning internal parameters could potentially enhance the performance of these sensors in bad weather situations. The findings of this research are important because most advanced detection algorithms used in autonomous driving heavily rely on undisturbed lidar data. By understanding the limitations and disturbances that occur when using lidar sensors in adverse weather, researchers can develop more robust algorithms that can accurately perceive the environment even in challenging conditions. This study highlights the importance of evaluating lidar sensor performance in adverse weather scenarios and provides insights into potential improvements through parameter tuning. By addressing these challenges, autonomous vehicles can achieve higher levels of safety and reliability regardless of environmental conditions.