Paul J. Tremont

Macrolevel Model Development for Safety Assessment of Road Network Structures

Traffic safety is beginning to receive increasing attention at the stage of transportation planning. Although the features of road networks are an essential aspect of transportation planning, studies of the safety effects of network patterns remain limited. In this study, macrolevel safety models were developed to explore the relationship between crash occurrence and several underlying variables, including demographic, land use, and road network variables. Different indices (e.g., the meshedness coefficient and closeness centrality) of network structures were developed to examine the effects of the network structure on safety at the zonal level. In many cases, a large percentage of the crash locations (especially for arterial crashes) was not related to the traffic analysis zones (TAZs) where drivers lived. To link crashes and features properly at the zonal level, crashes of each TAZ were modeled separately for non-state-maintained arterials (i.e., off-system roadways) and state-maintained arterials (i.e., on-system roadways). Several conditional autoregressive Bayesian models that incorporated the spatial correlation of nearby zones were developed. Estimation results showed that crashes occurring on non-state-maintained roads correlated more closely with the zonal network structure and the demographic characteristics inside the TAZ than did crashes occurring on state-maintained arterials; crashes on state-maintained arterials correlated more closely with the traffic and features of the major roads. The categorical variable generated from the meshedness coefficient captured well the nature of the relationship of network patterns to off-system road crashes. The results of this study indicate that both the roadway type and the structure of the road network should be considered when TAZ-level safety models are developed.

Safety modeling of urban arterials in Shanghai, China

Traffic safety on urban arterials is influenced by several key variables including geometric design features, land use, traffic volume, and travel speeds. This paper is an exploratory study of the relationship of these variables to safety. It uses a comparatively new method of measuring speeds by extracting GPS data from taxis operating on Shanghais urban network. This GPS derived speed data, hereafter called Floating Car Data (FCD) was used to calculate average speeds during peak and off-peak hours, and was acquired from samples of 15,000+ taxis traveling on 176 segments over 18 major arterials in central Shanghai. Geometric design features of these arterials and surrounding land use characteristics were obtained by field investigation, and crash data was obtained from police reports. Bayesian inference using four different models, Poisson-lognormal (PLN), PLN with Maximum Likelihood priors (PLN-ML), hierarchical PLN (HPLN), and HPLN with Maximum Likelihood priors (HPLN-ML), was used to estimate crash frequencies. Results showed the HPLN-ML models had the best goodness-of-fit and efficiency, and models with ML priors yielded estimates with the lowest standard errors. Crash frequencies increased with increases in traffic volume. Higher average speeds were associated with higher crash frequencies during peak periods, but not during off-peak periods. Several geometric design features including average segment length of arterial, number of lanes, presence of non-motorized lanes, number of access points, and commercial land use, were positively related to crash frequencies.

The influence of combined alignments on lateral acceleration on mountainous freeways: a driving simulator study

Combined horizontal and vertical alignments are frequently used in mountainous freeways in China; however, design guidelines that consider the safety impact of combined alignments are not currently available. Past field studies have provided some data on the relationship between road alignment and safety, but the effects of differing combined alignments on either lateral acceleration or safety have not systematically examined. The primary reason for this void in past research is that most of the prior studies used observational methods that did not permit control of the key variables. A controlled parametric study is needed that examines lateral acceleration as drivers adjust their speeds across a range of combined horizontal and vertical alignments. Such a study was conducted in Tongji Universitys eight-degree-of-freedom driving simulator by replicating the full range of combined alignments used on a mountainous freeway in China. Multiple linear regression models were developed to estimate the effects of the combined alignments on lateral acceleration. Based on these models, domains were calculated to illustrate the results and to assist engineers to design safer mountainous freeways.

Systematic Approach to Hazardous-Intersection Identification and Countermeasure Development

Safety performance functions (SPFs) are typically used to correlate geometric, traffic and environmental characteristics with total crashes and to identify hotspots which have high overall crash frequencies. However, with a distinct conflict pattern in vehicle maneuvers, each crash type is likely to associate with different risk factors. This study developed approach-level SPFs using a full Bayesian method to assess the safe effects of specific risk factors for rear-end, left-turn, right-angle, sideswipe and total crashes. To account for the spatial correlations among approaches at the same intersection, a random intersection-specific effect term was incorporated into each model. It was affirmed that these models were helpful in identifying high risk intersections with specific safety problems, and could serve as useful complements to general hotspot analyses using expected crash totals. In addition, it was found that certain variables (e.g. number of through lanes, median, and left-turn protection all on the entering approach) could have even contrary effects on crash occurrence of different types. Approach-level crash type models provide valuable insights in developing countermeasures aimed at reducing certain crash types and an improved ability in identifying deficiencies related to geometric and traffic characteristics for each intersection approach.