Developing Adaptive Border Crossing Mobility Measures and Short-Term Travel TIme Prediction Model Using Multiple Data Sets

Performance measures are critical for stakeholder's decision making; for instance, government agencies (e.g., state departments of transportation, metropolitan planning organization, U.S. Customs and Border Protection) need theme for infrastructure and mobility improvement. Similarly, road users such as shippers, commuters, and truckers need these performance measures to decide their route and departure time. For example, in the United States, the Highway Performance Monitoring System is a national highway information system that includes data on the extent, condition, performance, use, and operating characteristics of the nation's highways. Similarly, the Urban Mobility Scorecard lists various highway performance measures that are widely reported to compare urban mobility across the United States. However, these performance measures mostly cater to mobility on the highway system where the traffic conditions are completely different from that at the border crossing. Further, the current urban mobility performance measures do not translate well on the land port of entries (LPOEs) given the complexity of the border crossing operations including specific operation time, the number of inspection lanes open, inspection time, inspection type, approaching volume, commodity/shipment types, flow control and segmentation, and queue length. Another difference between the LPOEs and highways is that LPOEs are usually congested during operational hours, so defining a benchmark is difficult, unlike urban mobility performance measures where the baseline condition is free flow travel time. Current literature lacks border-specific mobility measures that can meet the need of various stakeholders interested in freight mobility at international LPOEs. The existing border performance measures such as wait time and crossing time are of less value since these do not provide historic trends and future trend, which are more valuable in short-term and long-term decision making. Moreover, there is no single performance measure that can be used for comparing multiple LPOEs and can adapt to multiple data sources. Recent Center for International Intelligent Transportation Research initiatives have provided access to real-time volumes at the LPOEs in El Paso, TX -- specifically at Zaragoza-Ysleta Bridge -- that is archived and aggregated by time intervals. The latest upgrade to the Border Crossing Information System (BCIS), which Texas A&M Transportation Institute operated and maintained, now provides the Free and Secure Trade (FAST) and non-FAST crossing times for U.S.-bound commercial vehicles. Despite the increased data availability at LPOEs, there is still a lack of meaningful performance for all stakeholders that can adapt to multiple data sets and can also be used for comparing LPOE performance. As part of this project, researchers investigated and developed the freight mobility performance measures that could present a realistic picture of border crossing performance. The research team explored various data sets that can be used for developing these performance measures with a focus on LPOEs in El Paso. Researchers developed new standard procedures for cleaning noise in the data sets, methodology for identifying the outliers, and working with the data sets. After developing standard methodology and cleaning the data, researchers developed a short-term wait time estimation algorithm. The team also performed a trend analysis using the cleaned data to develop historic trends. They then developed a methodology to calculate border performance scores in a consistent manner using one or more multiple data sets. These newly developed performance scores are independent of the variables that go into developing the performance measures, hence allowing a comparison among multiple LPOEs. These performance scores can combine multiple performance measures into one measure.