Pedestrian’s Safety and Mobility at Signalised Intersections, and Smart Crossing Decision System

Pedestrian safety becoming a serious issue, especially in developing nations, wherein higher crash rates have been reported by the World Health Organisation. Despite evidence suggesting lower pedestrian safety at signalised intersections in urban areas, there is a lack of in-depth investigation in most developing countries. Motivated by this need, the first contribution of this research is to present a novel approach that will help developing countries to determine and explain pedestrian crash causes while considering various challenges in these contexts. This was achieved by identifying the significant roadway environment characteristics influencing pedestrian vehicle crashes at signalised intersections in Amman, capital of Jordan. 166 accidents occurred during the period of 2007–2019 at 47 signalised intersections in Amman were analysed. The multilevel Generalised Linear Mixed Gamma regression (GLMG) model is the best fit for the data, indicating significant positive correlations between pedestrian crash frequencies and Annual Average Daily Traffic, pedestrian crossing volume, number of lanes, average lane width, and number of parking sides. Conversely, commercial land use and the presence of public transit facilities showed significant negative correlations with pedestrian crashes. The second contribution of this research is to establish a complementary approach that combines both observational behaviours (by using video recordings) and self-reported behaviours (by using a questionnaire) in order to explore the effects of multiple factors on pedestrian behaviour. This provides a deep understanding of pedestrians’ behaviour and differentiates their risky behaviours, which can aid in reducing the large number of pedestrian crashes and fatalities associated with pedestrians’ mistakes while walking. This was achieved by, firstly, analysing the observed pedestrian crossing behaviours in video records for the 47 signalised intersections based on the changes of pedestrians’ crossing time. The Generalised Linear Model (GLM) is the best fit for the data, indicating significant positive correlations between crossing time and carrying heavy bags, crossing in group, wet road surface, industrial land use, street width, average hourly vehicular volume, and duration of the ‘steady Don’t Walk phase’. Whereas males and rainy weather showed significant negative correlations with the crossing time. Secondly, a self-reporting pedestrian behaviour questionnaire for the Jordanian population (JPBQ) was developed and validated, spanning all ages; indeed, it represents the first valid long pedestrian behaviour questionnaire not only in Jordan but also in the Middle East, to the best of my iii knowledge, and can be used for purposes of pedestrian behaviour research. Additionally, this proposed JPBQ fills the gap found in previous studies in terms of the low reliability of the positive behaviour factor. Thus, developing this valid JPBQ is the third meaningful contribution made by this research. The JPBQ consisted of 40 items describing pedestrian behaviour, whilst the validation study itself included 400 participants. Principal component analysis (PCA) revealed a four-factor structure: transgressions, lapses, positive behaviours, and aggressive behaviours for both Long (31-item) and short (20 item) versions of the JPBQ, confirming validity and reliability for each factor. Across the four factors, the highest mean scores that pedestrians reported were for positive behaviours, while the least commonly reported behaviours were aggressive behaviours and lapses. Furthermore, the results of many statistical tests indicate significant behavioural differences amongst participants based on the effects of many variables such as the age, gender, marital status, and walking alternatives. Finally, a concept for a Smart Crossing Decision System (SCDS) that supports the crossing decisions of sighted pedestrians at signalised intersections is proposed alongside its architecture. As the existing SCDSs that have been designed worldwide across targeted blind and visually impaired pedestrians, with lack of such systems targeted sighted pedestrians. My system was designed based on the concept of context awareness that senses different navigation and traffic data, reasons about ability to cross streets safely, and reacts upon it by providing crossing decision to the pedestrian. Additionally, a novel algorithm for the crossing decision process is proposed. This algorithm considers many influencing factors (such as gender, waiting time, weather condition) of pedestrian crossing time which may have an important role in the crossing decision process when comparing the expected crossing time to the remaining green time of a pedestrian signal. This improves outcomes of the process of crossing decision making and adds another degree of certainty.