Times are displayed in (UTC-04:00) Eastern Time (US & Canada) Change
Towards multimodal freight network modeling: Geospatial map-matching of AIS data and integration with truck movements [Student Honor Award]
Automatic Identification System (AIS) consists of vessel’s traffic data, collected for navigational safety purposes (e.g., collision avoidance). The AIS data is continuous and ubiquitous over time and space, capturing timestamped locations of vessels. Historic AIS data can be leveraged for long-range freight planning purposes, project prioritization, etc. While previous studies successfully reconstructed vessel trajectories from AIS data, they were unable to assign reconstructed trip chains to a representative inland waterway network. The ability to map vessel trips to a waterway network, and for that network to connect to truck and rail networks, is critical for developing true multimodal freight Travel Demand Models (TDMs). This work improves upon existing methods to analyze trajectories from AIS data by developing and applying a geospatial map-matching algorithm to identify vessel trips, assign those trips to a representative inland waterway network, and generate multimodal freight flows integrating water and land side movements. To integrate waterway flows with land side movements, vessel trips identified by the algorithm are fused with truck trips (observed from GPS data) to and from freight port facilities, generating freight port “catchment areas” for project evaluation and prioritization. The methodology is applied to AIS data on the Arkansas River to calibrate and validate algorithm parameters. Multimodal integration results show that each port’s catchment area varies significantly, indicating that adopting an arbitrary radial impact area for different ports would lead to inaccurate project benefit estimates. Through the approaches introduced in this work, we highlight the value of AIS data for long-range freight planning and recommend future data elements to be included in AIS datasets. As we experience the rapid evolution of autonomous and automated technologies, we anticipate significant increases in the volume and veracity of vehicle and vessel movement data, giving our approach for trip identification and multimodal catchment area definition continued support.
About the Presenter

Magdalena Asborno
Senior Consultant
USACE-ERDC (contractor)
Magdalena Asborno graduated as a Civil and Transportation Engineer in 2003. After graduation, she got a Masters in Energy in Europe, and added value to construction projects of public works world-wide for more than a decade. In 2016, Magdalena joined the Freight Transportation Data Lab at the University of Arkansas, where she obtained a PhD degree in Civil Engineering, focused on freight transportation planning and systems. Dr. Asborno is currently a Senior Consultant for the US Army Corps of Engineers, Engineer Research and Development Center. At the Coastal and Hydraulics Laboratory, she applies research in multimodal data fusion and commodity-based network modeling, to support USACE’s Navigation Mission and better inform performance-based transportation infrastructure investment decision-making. Focusing on waterways, Dr. Asborno research interests include exploring the use of GIS tools, machine learning, and optimization models for multimodal “big data” analysis.
Presentation
Towards multimodal freight network modeling: Geospatial map-matching of AIS data and integration with truck movements [Student Honor Award]
Description