Asad J. Khattak

The objective of this work was to determine the effect of visibility-related factors and some environmental and human factors on the severity of pedestrian-vehicle crashes. It was hypothesized that decreasing visibility, contributed to by factors such as lighting, number of lanes, speed limit, and weather, are associated with an increase in injury severity. Some of the key results of the final model indicate that higher speed limits, less light conditions, and no traffic controls were significantly correlated with increased pedestrian injury severity when roadway visibility factors were under consideration. Dusk and dark with or without lighting were found to be factors correlated with increased pedestrian injury severity, while inclement weather was found to be correlated with lower pedestrian injury severity when environmental visibility-related factors were considered. Furthermore, a spatial autocorrelation revealed a high concentration of pedestrian–vehicle crashes in the Nashville and Memphis areas. This work is similar to prior works in their goal to study factors that affect pedestrian injury severity. While other models have looked at a large range of possible factors that may affect pedestrian injury severity, the model developed in this work focuses on visibility factors, environmental factors, and human-related factors. Another contribution is the data and modeling of the data. This study utilizes a dataset from Tennessee with more categories recorded for the visibility-related factors and applies a multinomial logistic regression model to the data.

The COVID-19 pandemic has highlighted the importance of information and communication technologies (ICTs) in providing virtual engagement. Planners and engineers must determine whether cities will see reductions in travel demand, given the increasing use of ICTs. Notably, ICTs facilitate online shopping and working from home (WFH). Generally, online shopping may lead to fewer shopping trips; similarly, WFH may reduce work-related trips. However, more WFH has the potential to generate other non-work trips, including shopping trips. To find answers and explore interdependencies, this study integrates pre-pandemic behavioral data with during-pandemic travel data. In our framework, WFH and online shopping are considered together. By harnessing the pre-pandemic 2017 National Household Travel Survey data, this study jointly analyzes the relationships between shopping trips, online shopping, and WFH with a conditional mixed process model that can address unobserved endogeneity and selection bias. The results suggest that, before the pandemic, online shopping was associated with fewer in-person shopping trips while WFH was associated with more shopping trips. The role of socio-demographic, locational, and travel-related factors is also explored. The during-pandemic data and analysis capture how COVID-19 affected travel behavior. Results show that the relationships among the key variables found in the pre-pandemic data are similar but differ in magnitude from during the pandemic. WFH increased from 12% to 61% during COVID-19, admittedly an unusual situation. In the next “new normal,” planners may improve travel demand models by treating WFH explicitly as an alternative to traveling to work in the trip generation and time of day models.