The significance of endogeneity problems in crash models: An examination of left-turn lanes in intersection crash models

Crash prediction models are used for a variety of purposes including forecasting the expected future performance of various transportation system segments with similar traits. The influence of intersection features on safety have been examined extensively because intersections experience a relatively large proportion of motor vehicle conflicts and crashes compared to other segments in the transportation system. The effects of left-turn lanes at intersections in particular have seen mixed results in the literature. Some researchers have found that left-turn lanes are beneficial to safety while others have reported detrimental effects on safety. This inconsistency is not surprising given that the installation of left-turn lanes is often endogenous, that is, influenced by crash counts and/or traffic volumes. Endogeneity creates problems in econometric and statistical models and is likely to account for the inconsistencies reported in the literature. This paper reports on a limited-information maximum likelihood (LIML) estimation approach to compensate for endogeneity between left-turn lane presence and angle crashes. The effects of endogeneity are mitigated using the approach, revealing the unbiased effect of left-turn lanes on crash frequency for a dataset of Georgia intersections. The research shows that without accounting for endogeneity, left-turn lanes ‘appear’ to contribute to crashes; however, when endogeneity is accounted for in the model, left-turn lanes reduce angle crash frequencies as expected by engineering judgment. Other endogenous variables may lurk in crash models as well, suggesting that the method may be used to correct simultaneity problems with other variables and in other transportation modeling contexts.

Applying Bayesian hierarchical models to examine motorcycle crashes at signalized intersections

Motorcycles are overrepresented in road traffic crashes and particularly vulnerable at signalized intersections. The objective of this study is to identify causal factors affecting the motorcycle crashes at both four-legged and T signalized intersections. Treating the data in time-series cross-section panels, this study explores different Hierarchical Poisson models and found that the model allowing autoregressive lag 1 dependent specification in the error term is the most suitable. Results show that the number of lanes at the four-legged signalized intersections significantly increases motorcycle crashes largely because of the higher exposure resulting from higher motorcycle accumulation at the stop line. Furthermore, the presence of a wide median and an uncontrolled left-turn lane at major roadways of four-legged intersections exacerbate this potential hazard. For T signalized intersections, the presence of exclusive right-turn lane at both major and minor roadways and an uncontrolled left-turn lane at major roadways of T intersections increases motorcycle crashes. Motorcycle crashes increase on high-speed roadways because they are more vulnerable and less likely to react in time during conflicts. The presence of red light cameras reduces motorcycle crashes significantly for both four-legged and T intersections. With the red-light camera, motorcycles are less exposed to conflicts because it is observed that they are more disciplined in queuing at the stop line and less likely to jump start at the start of green.

Factors affecting motorcycle crashes at signalized intersections

Singapore crash statistics show that motorcycles are involved in about 54% of crashes at intersections. Moreover, about 46% of fatal and 67% of injury motorcycle crashes occur at signalized intersections. The objective of this study is to identify causal factors affecting the motorcycle crashes at both four-legged and three-legged signalized intersections. Treating the data in time-series cross-section panels, this study explores different Hierarchical Poisson models and found that the model allowing autoregressive lag 1 dependent specification in the error term is the most suitable. Analysis of the results shows the number of lanes at the intersections significantly increases motorcycle crashes largely because of the higher exposure resulting from higher motorcycle accumulation at the stop line. Furthermore, the presence of a wide median at four-legged intersections and an exclusive right-turn lane and an uncontrolled left-turn lane at three-legged intersections exacerbate this potential hazard. Moreover, motorcycle crashes increase on high-speed roadways because of the vulnerability of the motorcyclists. The presence of red light cameras reduces motorcycle crashes significantly on the intersection roadways for both four-legged and three-legged intersections. With the red-light camera, motorcycles are less exposed to conflicts because it is observed that they are more disciplined in queuing at the stop line and less likely to jump start at the start of green.

Fusion of in-vehicle sensor data to develop Intelligent Driver Training System (IDTS)

The over represented number of novice drivers involved in crashes is alarming. Driver training is one of the interventions aimed at mitigating the number of crashes that involve young drivers. To our knowledge, Advanced Driver Assistance Systems (ADAS) have never been comprehensively used in designing an intelligent driver training system. Currently, there is a need to develop and evaluate ADAS that could assess driving competencies. The aim is to develop an unsupervised system called Intelligent Driver Training System (IDTS) that analyzes crash risks in a given driving situation. In order to design a comprehensive IDTS, data is collected from the Driver, Vehicle and Environment (DVE), synchronized and analyzed. The first implementation phase of this intelligent driver training system deals with synchronizing multiple variables acquired from DVE. RTMaps is used to collect and synchronize data like GPS, vehicle dynamics and driver head movement. After the data synchronization, maneuvers are segmented out as right turn, left turn and overtake. Each maneuver is composed of several individual tasks that are necessary to be performed in a sequential manner. This paper focuses on turn maneuvers. Some of the tasks required in the analysis of ‘turn’ maneuver are: detect the start and end of the turn, detect the indicator status change, check if the indicator was turned on within a safe distance and check the lane keeping during the turn maneuver. This paper proposes a fusion and analysis of heterogeneous data, mainly involved in driving, to determine the risk factor of particular maneuvers within the drive. It also explains the segmentation and risk analysis of the turn maneuver in a drive.

The impact of red light cameras on safety in Arizona

Red light cameras (RLCs) have been used in a number of US cities to yield a demonstrable reduction in red light violations; however, evaluating their impact on safety (crashes) has been relatively more difficult. Accurately estimating the safety impacts of RLCs is challenging for several reasons. First, many safety related factors are uncontrolled and/or confounded during the periods of observation. Second, “spillover” effects caused by drivers reacting to non-RLC equipped intersections and approaches can make the selection of comparison sites difficult. Third, sites selected for RLC installation may not be selected randomly, and as a result may suffer from the regression to the mean bias. Finally, crash severity and resulting costs need to be considered in order to fully understand the safety impacts of RLCs. Recognizing these challenges, a study was conducted to estimate the safety impacts of RLCs on traffic crashes at signalized intersections in the cities of Phoenix and Scottsdale, Arizona. Twenty-four RLC equipped intersections in both cities are examined in detail and conclusions are drawn. Four different evaluation methodologies were employed to cope with the technical challenges described in this paper and to assess the sensitivity of results based on analytical assumptions. The evaluation results indicated that both Phoenix and Scottsdale are operating cost-effective installations of RLCs: however, the variability in RLC effectiveness within jurisdictions is larger in Phoenix. Consistent with findings in other regions, angle and left-turn crashes are reduced in general, while rear-end crashes tend to increase as a result of RLCs.

Examining Exposure of Motorcycles at Signalized Intersections

Crash statistics in Singapore from 2001 to 2005 have shown that motorcycles are involved in about 54% of intersection crashes. The overall involvement of motorcycles in crashes as the not-at-fault party is about 43% but at intersections, the corresponding percentage is increased to 57%. Quasi-induced exposure estimates show that the motorcycle exposure rate at signalized intersections is 41.7% even though motorcycles account for only 19% of the vehicle population. This study seeks to examine in greater details, the problem of motorcycle exposure at signalized intersections. In particular, the exposure arising from potential crashes with red light running vehicles from the conflicting stream at four signalized intersections is investigated. The results show that motorcycles are more exposed because they tend to accumulate near the stop-line during the red phase to facilitate an earlier discharge during the initial period of the green which is the more vulnerable period. At sites where there are more weaving opportunities because the lanes are wider or where there are exclusive right-turn lanes, the accumulation is higher and hence an increased exposure is observed. The analysis also shows that the presence of heavy vehicles tends to decrease motorcycle exposure as their weaving opportunities become restricted as well as there is a greater reluctance for them to weave past or queue alongside the heavy vehicles and their effects intensify for narrower lane width.

Effect of upstream traffic signal on capacity of a downstream TWSC intersection

This paper investigates the platoon dispersion model that is part of the 2010 Highway Capacity Manual that is used for forecasting downstream traffic flows for analyzing both signalized and TWSC intersections. The paper focuses on the effect of platoon dispersion on the proportion of time blocked, the conflicting flow rate, and the capacity flow rate for the major street left turn movement at a TWSC intersection. The existing HCM 2010 methodology shows little effect on conflicting flow or capacity for various distances downstream from the signalized intersection. Two methods are suggested for computing the conflicting flow and capacity of minor stream movements at the TWSC intersection that have more desirable properties than the existing HCM method. Further, if the existing HCM method is retained, the results suggest that the upstream signals model be dropped from the HCM method for TWSC intersections.

Friedman Economic Policies, Social Exclusion and Crime: Toward a Gendered Left Realist Subcultural Theory.

Left realists contend that people lacking legitimate means of solving the problem of relative deprivation may come into contact with other frustrated disenfranchised people and form subcultures, which in turn, encourage criminal behaviors. Absent from this theory is an attempt to address how, today, subcultural development in North America and elsewhere is heavily shaped simultaneously by the recent destructive consequences of right-wing Friedman or Chicago School economic policies and marginalized men's attempts to live up to the principles of hegemonic masculinity. The purpose of this paper, then, is to offer a new left realist theory that emphasizes the contribution of these two key determinants.

Qualitative spatial reasoning about directions in a corridor containing a single turn

An experiment was conducted to investigate the process of reasoning about directions in an egocentric space. Each participant walked through a corridor containing an angular turn ranging in size from 0° to 90°, in 15° increments. A direction was given to participants at the entrance of the corridor and they were asked to answer this direction at the end of this corridor. Considering the fact that participants had to reason the direction in the featureless corridor, two hypotheses were proposed: (i) reasoning about directions falls into qualitative reasoning by using a small number of coarse angular categories (four 90° categories or eight 45° categories: 90° categories consist of front, back, left, right; 45° categories consist of 90° categories and the four intermediates) that reference axes generate; (ii) reasoning about directions would be done by recalling the rotation angle from the traveling direction to the direction that participants tried to answer. In addition, the configuration of reference axes that participants employed was examined. Both hypotheses were supported, and the data designated that reference axes consisted of eight directions: a pair of orthogonal axes and diagonals.

Woodturning for older learners : a real 'turn on'

The research reported in this study concerns older adults from Australia who voluntarily chose to learn the craft of woodturning. Semi-structured interviews and a survey questionnaire were distributed to members of a woodturning club to explore their motivations and the processes by which they learned how to woodturn. The findings indicated that participants’ motivation could be construed as both intrinsic and extrinsic. They used seven approaches to learning – structured courses, instruction from convenors, modelling/watching/demonstrations, guided practice and monitoring by convenors, trial and error with practice, advice and help from peers and reading. Finally, the positive climate of the organisation was found to be particularly important to the older learners.

The influence of oblique viewing on axial and peripheral refraction for emmetropes and myopes

Refraction may be affected by the forces of lids and extraocular muscles when eye direction and head direction are not aligned (oblique viewing) which might potentially influence past findings on peripheral refraction of the eye. We investigated the effect of oblique viewing on axial and peripheral refraction. In a first experiment, cycloplegic axial refractions were determined when subjects' heads were positioned to look straight-ahead through an open-view autorefractor and when the heads were rotated to the right or left by 30° with compensatory eye rotation (oblique viewing). Subjects were 16 young emmetropes (18–35 years), 22 young myopes (19–36 years) and 15 old emmetropes (45–60 years). In a second experiment, cycloplegic peripheral refraction measurements were taken out to ±34° horizontally from fixation while the subjects rotated their heads to match the peripheral refraction angles (eye in primary position with respect to the head) or the eyes were rotated with respect to the head (oblique viewing). Subjects were 10 emmetropes and 10 myopes. We did not find any significant changes in axial or peripheral refraction upon oblique viewing for any of the subject groups. In general for the range of horizontal angles used, it is not critical whether or not the eye is rotated with respect to the head during axial or peripheral refraction.

Extraction of road lanes from high-resolution stereo aerial imagery based on maximum likelihood segmentation and texture enhancement

Accurate road lane information is crucial for advanced vehicle navigation and safety applications. With the increasing of very high resolution (VHR) imagery of astonishing quality provided by digital airborne sources, it will greatly facilitate the data acquisition and also significantly reduce the cost of data collection and updates if the road details can be automatically extracted from the aerial images. In this paper, we proposed an effective approach to detect road lanes from aerial images with employment of the image analysis procedures. This algorithm starts with constructing the (Digital Surface Model) DSM and true orthophotos from the stereo images. Next, a maximum likelihood clustering algorithm is used to separate road from other ground objects. After the detection of road surface, the road traffic and lane lines are further detected using texture enhancement and morphological operations. Finally, the generated road network is evaluated to test the performance of the proposed approach, in which the datasets provided by Queensland department of Main Roads are used. The experiment result proves the effectiveness of our approach.