Moshe E. Ben-Akiva, Vittorio Marzano, Fang Zhao, Jinping (Jenna) Guan
The Federal Highway Administration of the United States Department of Transportation
Moshe E. Ben-Akiva, Chris Zegras, Vittorio Marzano (University of Napoli Federico II), Fang Zhao (Singapore-MIT Alliance for Research and Technology), Kyungsoo Jeong, Jinping (Jenna) Guan, Peiyu Jing, William Wong
- Development of a system for tracking vehicles using GPS loggers, tablets, and post-processing of raw data (i.e., stop detection, map matching, machine learning based on context data and user history). This system will be based on the Future Mobility Sensing (FMS) technology, which has been developed and tested for passenger travel surveys.
- Adaptation of vehicle tracking algorithms and technologies for shipment tracking while accounting for specific issues, e.g., sampling frequency.
- Adaptation of the map matching and stop detection algorithms for shipment tracking while accounting for specific issues, e.g., Point of Interest (POI) data.
- Design of a web-based interface for verification of tracking data that is tailored to each type of respondent and embedded with data visualization features. Implementation of processed sensing data verification based on interactions with respondents through innovative survey capabilities.
- Pilot study implementation to be conducted in both intercity and urban environments. Different functionalities will be tested in an integrated or a modular fashion, which include vehicle tracking, shipment tracking, and truck driver surveys.
This project aims at developing an integrated approach for future freight and logistics surveys including all relevant freight entities: establishments (including logistics operators and 3PL), carriers/drivers/vehicles, and shipments. The proposed approach leverages a coherent and holistic survey methodology and fully integrated survey instruments. It is based on innovative and scalable technologies with considerable time and geographical coverage (national, regional/urban and rural areas).
The key concept underlying the research is the extension to freight of the Future Mobility Survey (FMS) tool, which has already been proven effective in passenger surveys. FMS makes use of smartphones/tablets and GPS loggers, advanced sensing and communication technologies and machine learning algorithms to collect data reflecting what all relevant freight agents do, not what they say they do. State-of-the-art sensing devices enhance the quality and quantity of data, especially when combined with information from agents themselves.
The framework is, and it aims to ameliorate inherent limitations in current freight data collection methods, obtain unprecedented freight data for statistical purposes, and enable the implementation of a new generation of freight models, including agent-based models.
The 3-year project is articulated into four main project tasks: 1. Innovative tracking of vehicles and shipments; 2. Development of innovative sensing-based survey and visualization tools; 3. Integration of FMS with data and modelling needs in freight; 4. Case studies.
The vision of this project is to create a new paradigm for freight surveys that leverages state-of-the-art technologies. This project will utilize vehicle tracking devices such as GPS loggers and smartphone/tablet applications to collect raw data regarding freight activities in both urban and intercity environments.
FMS technology will be utilized to combine the valuable information from state-of-the-art sensing devices with prompted-recall surveys from respondents. While tracking and sensing techniques provide accurate information on routes and stops, such data do not provide enough information on explanatory factors such as decision-maker characteristics and transportation attributes, e.g., shipment value and cost. However, these factors are necessary for behavioral modeling. Thus such additional information will be collected in daily surveys to gather further details regarding driver decisions. This is being done through web-based verification of tracking data. In addition, user-specific functionalities and visualizations will be embedded into the FMS technology.
Two pilot case studies will be conducted with the developed data collection techniques. The case studies will include integrated analysis of urban and intercity freight in the US. A proof of concept will illustrate FMS capabilities in addressing challenges existing in traditional freight data collection. The case study will also look to provide a better understanding of truck driver behavior. The final result of the case studies will be analyzed and findings will be documented.
This developed FMS tool will be used to enhance freight survey and modeling techniques in the US. This includes analyzing opportunities and challenges in surveying freight and logistics with the FMS tool, analyzing new opportunities for survey recruitment and collaboration, and developing guidelines on how FMS will enable the implementation of a new generation of freight models, e.g., agent-based behavioral models.
- Integrated CFS, “Integrated Freight Survey, Shipment Tracking, and Vehicle Tracking”, Innovations in Freight Data Workshop (TRB), Irvine, CA, May 2017
- “Combing gps data collection with assisted machine learning to enhance freight vehicle and driver surveys: methodology and demonstration”, submissions to TRB 2018 for presentation and journal for publication, August 2017
- “Intercity Truck Driver Route Choice Modeling Capturing Incorporating Drivers’ Heterogeneity In Predicating Use Of Toll Road Usage: Data Collection, Model Estimation, And Model Application”, submissions to TRB 2018 for presentation, August 2017
- Future Freight and Logistics Survey (shipment survey), accepted for the 2017 METRANS International Urban Freight Conference (I-NUF), October 2017