Observing Traffic -- Utilizing a Ground Based LiDAR and Observation Protocols at a T-Junction in Germany

Introducing automated vehicles onto current urban traffic conditions is a demanding task. While trajectory and maneuver planning algorithms needs to cope with complex road structures, the automation needs to deal with another volatile element – human road users. Understanding how traffic participants behave nowadays is indispensable to ensure that automated vehicles can react appropriately to other road users. This paper aims to provide an overview of observation methodologies capable of quantifying different aspects of road user behavior. An observation study of a T-junction in German rush-hour traffic using a ground based LiDAR and an HTML app for manual observations is presented. In normal traffic conditions drivers follows their road prioritization quite strictly. Once congestions emerge on the main road this behavior changes: prioritized drivers yield their right of way to let other vehicles turn onto the main road by increasing gap sizes, sometimes accompanied by flashing their headlights to inform the turning driver of the yielding behavior. Overall, the method of using a ground based LiDAR in combination with human observers has a high potential to quantitatively describe perceivable traffic occurrences while complying with data privacy laws.