Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
This paper introduces a real-time roadwork detection system for autonomous vehicles, fusing YOLO-based object detection with LiDAR data to identify and localize construction sites. The system merges individual roadwork objects into coherent sites and records their outlines in world coordinates, achieving sub-0.5m localization accuracy in real-world tests. Training leveraged both an adapted US dataset and a novel dataset collected in Berlin.
Autonomous vehicles can now navigate chaotic roadwork zones with sub-meter accuracy thanks to a new real-time detection system.
Road construction sites create major challenges for both autonomous vehicles and human drivers due to their highly dynamic and heterogeneous nature. This paper presents a real-time system that detects and localizes roadworks by combining a YOLO neural network with LiDAR data. The system identifies individual roadwork objects while driving, merges them into coherent construction sites and records their outlines in world coordinates. The model training was based on an adapted US dataset and a new dataset collected from test drives with a prototype vehicle in Berlin, Germany. Evaluations on real-world road construction sites showed a localization accuracy below 0.5 m. The system can support traffic authorities with up-to-date roadwork data and could enable autonomous vehicles to navigate construction sites more safely in the future.