Using turn signals to convey a driver's intention to change lanes provides a direct and unambiguous way of communicating with nearby drivers. Nonetheless, past research has indicated that drivers may not always use their turn signals prior to starting a lane change. In this study, we analyze realistic driving data to investigate turn signal usage during lane changes on highways in and around Gothenburg, Sweden. We examine turn signal usage and identify factors that have an influence on it by employing Bayesian hierarchical modelling (BHM). The results showed that a turn signal was used in approximately 60% of cases before starting a lane change, while it was only used after the start of a lane change in 33% of cases. In 7% of cases, a turn signal was not used at all. Additionally, the BHM results reveal that various factors influence turn signal usage. The study concludes that understanding the factors that affect turn signal usage is crucial for improving traffic safety through policy-making and designing algorithms for autonomous vehicles for future mixed traffic.

翻译：使用转向灯向附近驾驶员传达变道意图，是一种直接且明确的沟通方式。然而，以往研究表明，驾驶员在开始变道前并不总是使用转向灯。本研究通过分析瑞典哥德堡及其周边高速公路上的真实驾驶数据，探究车道变换过程中转向灯的使用情况。我们采用贝叶斯分层建模（BHM）考察转向灯的使用行为，并识别影响该行为的关键因素。结果显示：约60%的车道变换在开始前使用了转向灯，33%在变道开始后才使用，而7%的变道全程未使用转向灯。此外，BHM分析揭示多种因素共同影响转向灯使用。研究结论指出，理解影响转向灯使用的因素对于通过政策制定提升交通安全，以及为未来混合交通设计自动驾驶算法具有重要意义。