Salzburg Intersection Trajectory and PET Interaction Dataset (SIT-PET)

This dataset provides high-resolution road user trajectory data and PET-based interaction events collected at a signalized urban intersection in Salzburg, Austria. Trajectories of vehicles, pedestrians, and two-wheelers are complemented by predefined interaction scenarios and corresponding Post-Encroachment Time (PET) computations, enabling detailed analysis of spatiotemporal interactions at the intersection. The dataset is intended for research on surrogate safety measures, traffic behavior analysis, and method development in traffic safety studies. PET events are provided as raw computation results without criticality filtering, allowing users to apply their own thresholds and definitions. Static object attributes, scenario definitions, traffic signal states, and weather observations are included to support contextual analyses. Due to data volume constraints, the published dataset represents a six-month representative test subset of a larger dataset while preserving the full structure and semantics. Detailed documentation of all tables, attributes, and relationships is provided in the accompanying README and schema files. An accompanying data descriptor article describing the data preparation in more detail was published in Data. Please cite the article if you use this dataset. Steinmaßl, M.; Rehrl, K.; Vornberger, T. SIT-PET: Long-Term Multimodal Traffic Trajectory Data with PET-Based Interaction Events at a Signalized Intersection. Data 2026, 11, 68. https://doi.org/10.3390/data11040068本数据集采集自奥地利萨尔茨堡一处城市信号控制交叉口，包含高分辨率道路使用者轨迹数据与基于碰撞后侵入时间（Post-Encroachment Time, PET）的交互事件数据。数据集涵盖机动车、行人及两轮车的轨迹数据，并补充了预定义交互场景与对应的PET计算结果，可支持对交叉口时空交互行为的精细化分析。 本数据集面向交通安全研究中的替代安全指标、交通行为分析以及方法开发相关研究。其中PET事件以原始计算结果形式提供，未进行临界性过滤，使用者可自行设定阈值与判定标准。此外数据集还包含静态地物属性、场景定义、交通信号状态与气象观测数据，以支撑上下文关联分析。 受限于数据体量，本次发布的数据集仅为全量数据集的6个月代表性测试子集，但完整保留了原数据集的全部结构与语义信息。所有数据表、属性字段及关联关系的详细文档均随附于README文件与模式文件中。 另有一篇配套的数据描述文章详细阐述了本数据集的制备流程，已发表于《Data》期刊。若您使用本数据集，请引用该文章： Steinmaßl, M.; Rehrl, K.; Vornberger, T. SIT-PET: Long-Term Multimodal Traffic Trajectory Data with PET-Based Interaction Events at a Signalized Intersection. Data 2026, 11, 68. https://doi.org/10.3390/data11040068