Διάλεξη του Καθηγητή Peter Savolainen του Πανεπιστημίου του Μίτσιγκαν στο ΕΜΠ – 11 Νοεμβρίου 2024

Stopping sight distance (SSD) has long been a concern among design engineers, given difficulties in attaining minimum values in certain roadway contexts in consideration of cost constraints. For example, SSD is one of the ten controlling criteria for design and documentation of design exceptions. However, the extant research literature has generally not shown that locations with insufficient SSD experience higher crash risks as compared to locations that meet or exceed recommended design values. This study investigates the relationships between SSD and crash risk using high-fidelity LiDAR data. These data are used to examine crash trends on the roadway network maintained by the Utah Department of Transportation. LiDAR data are used to assess compliance with the current SSD design policy as articulated in A Policy on the Geometric Design of Streets and Highways (i.e., the AASHTO Green Book). These data are integrated with crash and roadway geometric information. A series of count data models are estimated to assess the relationship between available SSD and the frequency of crashes while controlling for other important variables of interest. The results show that roadways with limited sight distance tended to experience significantly more crashes as compared to other similar segments where sight restrictions were not present. The findings from this study provide an empirical basis for estimating the potential impacts of design scenarios where it may be impractical to satisfy the minimum recommendation distance from the AASHTO Green Book. It can also help to inform agency practices and the development of projects to mitigate the sources of the sight distance limitations. Subsequent follow-up research in Michigan will also be briefly discussed.