01697481

Component

With new technologies, while the data availability at land port of entries (LPOEs) has increased, there is still a lack of predictive performance measures for meaningful use by stakeholders. For instance, instantaneous performance measures are after-the-fact with limited use for most stakeholders in terms of pro-active decision making. Therefore, as part of this study, the authors investigated new data sources for calculating border crossings performance measures, developed standard data cleaning procedures for these datasets and explored machine learning for prediction of short-term wait time at a US-Mexico border crossing. This research helps to better understand the performance of the LPOEs and predict situations when the performance deteriorates. This research will assist transportation and border protection agencies to be better prepared for making operations and mobility related decisions, while simultaneously helping truckers and shippers make real-time routing decisions.

This paper was sponsored by TRB committee AT020 Standing Committee on International Trade and Transportation.

19-05740

English

6p

2019

Transportation Research Board 98th Annual Meeting

Digital/other

Figures; References

Border crossings; Freight transportation; Machine learning; Mathematical prediction; Performance measurement; Trucking; Waiting time

United States-Mexico Border

Data and Information Technology; Freight Transportation; Operations and Traffic Management; Planning and Forecasting

Transportation Research Board Annual Meeting 2019 Paper #19-05740

TRIS, TRB, ATRI

Dec 7 2018 9:29AM