How big data serves for freight safety management at highway-rail grade crossings? A spatial approach fused with path analysis

详细信息    查看全文

• 作者：Jun Liu^a ; ^{jliu34@utk.edu" class="auth_mail" title="E-mail the corresponding author}Author Vitae ; Xin Wang^b ; ^{xin.wang@vdot.viginia.gov" class="auth_mail" title="E-mail the corresponding author}Author Vitae ; Asad J. Khattak^c ; ^{akattak@utk.edu" class="auth_mail" title="E-mail the corresponding author}Author Vitae ; Jia Hu^d ; ^{Jia.hu@dot.gov" class="auth_mail" title="E-mail the corresponding author}Author Vitae ; JianXun Cui^e ; ^{cuijianxun@hit.edu.cn" class="auth_mail" title="E-mail the corresponding author}Author Vitae ; Jiaqi Ma^f ; ^{jiaqi.ma@leidos.com" class="auth_mail" title="E-mail the corresponding author}Author Vitae
• 关键词：Big data ; Highway-rail grade crossing ; Truck crash ; Driving behavior ; Geographically ; weighted regressionPath analysis
• 刊名：Neurocomputing
• 出版年：2016
• 出版时间：12 March 2016
• 年：2016
• 卷：181
• 期：Complete
• 页码：38-52
• 全文大小：2442 K

文摘

Future vehicle warning systems needs a local (instead of global) analysis of real-time information transmitted between vehicles and infrastructures, to provide local warning information matching the instantaneous driving contexts. Spatial modeling techniques extracting the location information into the analysis fulfill the needs of local analysis. For truck-involved collisions at highway-rail crossings, the local warnings seem to matters more, than for traffic crashes at the normal highway segments and regular intersections. Crashes at rail grade crossings can result in severe injuries and fatalities to vehicle occupants, while truck-involved crashes at crossings can further result in serious damage to train, crossing and railway equipment. Truck-involved crashes at grade crossings have received limited attention compared with crashes involved with passenger cars. This study presents a methodology of improving safety of trucks at railroad crossings, by taking advantage of the big data containing location information. Unlike previous studies that constructed a direct relationship between the safety outcomes and associated factors, this study investigates direct relationships together with indirect relationships through the truck driver behaviors before collisions, using path analysis techniques. To sum up, this study applies a spatial approach fused with path analysis to uncover the local relationships between truck driver injury severity and crossing controls across the space. By doing so, the research is able to: i) provide a benchmark of identifying potentially risky vehicles on a real time basis; ii) evaluate current control devices at railroad crossing across the country and pin point the potentially problematic crossing sites. An empirical study was conducted by using a rich crash database from the Federal Railroad Administration (N= 4738 for 2004–2014). The results show that truck-involved crashes occurring at crossings without gate controls are generally associated with higher chance of injury, while the associations vary significantly across the space. In general, crashes in the Midwest and Great Lake regions are associated with an even higher chance of injury at crossings without gates, compared with other regions. More Results and implications are discussed in the paper.