Jiaxi Zhao
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Jiaxi Zhao has a total of 13 co-authors. Here are the top co-authors by number of shared publications:
- C. Tyler Dick
Univ. of Texas at Austin
6 shared publications - Di Kang
Department of Civil and Environmental Engineering, Rutgers, The State University of New Jersey, Piscataway, NJ
2 shared publications - Steven W. Kirkpatrick
Applied Research Associates, Inc., Los Altos, CA
2 shared publications - Xiang Liu
Department of Civil and Environmental Engineering, Rutgers, The State University of New Jersey, Piscataway, NJ
2 shared publications - Bojian Zhang1 shared publication
C. Tyler Dick
Department of Civil Architectural and Environmental Engineering University of Texas at Austin Austin Texas USA
1 shared publication
Latest Publications
- Open Access
- 18 October 2024
- Published by Institution of Engineering and Technology (IET) in IET Intelligent Transport Systems
- Vol. 18 (12) , 2359-2374
Enhancing freight train delay prediction with simulation‐assisted machine learning
Boosting the rail freight modal share is an ambitious target in Europe and North America. Yards, where freight trains are arranged, can be crucial in realizing this target by reliable dispatching to the network. This paper predicts freight train departures by developing a simulation‐assisted machine learning model with two concepts: general (adding all predictors at once) and step‐wise (adding predictors as they become available in sub‐yard operations) for hump yards with the conventional layout to provide a generalized model for European and North American contexts. The developed model is a decision tree algorithm, validated via 10‐fold cross‐validation. The model's performance on three data sets—a real‐world European yard, a baseline simulation, and an ultimate randomness simulation for a comparable North American yard—shows a respective of 0.90, 0.87, and 0.70. Step‐wise inclusion of the predictors results differently for the real‐world and simulation data. The global feature importance highlights maximum planned length, departure weekday, the number of arriving trains, and minimum arrival deviation as key predictors for the real‐world data. For the simulation data, the most significant predictors are departure yard predictors, the number of arriving trains, and the maximum hump duration. Additionally, utilization rates—except for the receiving yard—enhance the predictions.
- 18 October 2024
- Published by Institution of Engineering and Technology (IET) in IET Intelligent Transport Systems
- Vol. 18 (12) , 2359-2374
Open AccessEnhancing freight train delay prediction with simulation‐assisted machine learning
Boosting the rail freight modal share is an ambitious target in Europe and North America. Yards, where freight trains are arranged, can be crucial in realizing this target by reliable dispatching to the network. This paper predicts freight train departures by developing a simulation‐assisted machine learning model with two concepts: general (adding all predictors at once) and step‐wise (adding predictors as they become available in sub‐yard operations) for hump yards with the conventional layout to provide a generalized model for European and North American contexts. The developed model is a decision tree algorithm, validated via 10‐fold cross‐validation. The model's performance on three data sets—a real‐world European yard, a baseline simulation, and an ultimate randomness simulation for a comparable North American yard—shows a respective of 0.90, 0.87, and 0.70. Step‐wise inclusion of the predictors results differently for the real‐world and simulation data. The global feature importance highlights maximum planned length, departure weekday, the number of arriving trains, and minimum arrival deviation as key predictors for the real‐world data. For the simulation data, the most significant predictors are departure yard predictors, the number of arriving trains, and the maximum hump duration. Additionally, utilization rates—except for the receiving yard—enhance the predictions.
- 27 September 2024
- Published by SAGE Publications in Transportation Research Record: Journal of the Transportation Research Board
- Vol. 2679 (2) , 2013-2030
Quantifying the Influence of Tank Car Position and Train Configuration on the Risk of Rail Transport of Class 3 Flammable Liquids
As the amount of high-hazardous materials (hazmat) being shipped in high-hazard flammable unit trains (HHFUT) increases, mitigating the risks associated with railway transportation of hazmat remains an industry priority. Compared with HHFUT configurations, the placement of hazmat railcars in high-hazard flammable trains (HHFT) and the number of classification yards they traverse can affect the risk of derailment and subsequent hazmat release for HHFTs. This paper evaluates hazmat release risks associated with the transportation of U.S. Department of Transportation (U.S. DOT) Class 3 flammable liquids by tank cars. Specifically, it compares the risks of HHFUTs and HHFTs for shipping a designated amount of hazmat and considers both mainline and yard operations in alignment with existing regulations and practical guidance. The methodology quantifies the transportation risk in total expected release consequence (i.e., casualties) given the total amount of Class 3 flammable liquids transported in HHFUT and HHFT configurations with different tank car placement strategies. Based on the case study, we find that using five 100-railcar HHFTs, each with 20 tank cars in positions 66 through 85, is both practical and generates the lowest transportation risk. Some hazmat railcar placement strategies in HHFTs lead to a higher release risk than HHFUT operations, while others result in a lower release risk. The proposed methodology could be extended to diverse operating scenarios to better understand the impacts of train configuration and tank car placement on the risk of rail transport for flammable liquids.
- 27 September 2024
- Published by SAGE Publications in Transportation Research Record: Journal of the Transportation Research Board
- Vol. 2679 (2) , 2013-2030
Quantifying the Influence of Tank Car Position and Train Configuration on the Risk of Rail Transport of Class 3 Flammable Liquids
As the amount of high-hazardous materials (hazmat) being shipped in high-hazard flammable unit trains (HHFUT) increases, mitigating the risks associated with railway transportation of hazmat remains an industry priority. Compared with HHFUT configurations, the placement of hazmat railcars in high-hazard flammable trains (HHFT) and the number of classification yards they traverse can affect the risk of derailment and subsequent hazmat release for HHFTs. This paper evaluates hazmat release risks associated with the transportation of U.S. Department of Transportation (U.S. DOT) Class 3 flammable liquids by tank cars. Specifically, it compares the risks of HHFUTs and HHFTs for shipping a designated amount of hazmat and considers both mainline and yard operations in alignment with existing regulations and practical guidance. The methodology quantifies the transportation risk in total expected release consequence (i.e., casualties) given the total amount of Class 3 flammable liquids transported in HHFUT and HHFT configurations with different tank car placement strategies. Based on the case study, we find that using five 100-railcar HHFTs, each with 20 tank cars in positions 66 through 85, is both practical and generates the lowest transportation risk. Some hazmat railcar placement strategies in HHFTs lead to a higher release risk than HHFUT operations, while others result in a lower release risk. The proposed methodology could be extended to diverse operating scenarios to better understand the impacts of train configuration and tank car placement on the risk of rail transport for flammable liquids.
- 14 February 2024
- Published by Elsevier in Journal of Rail Transport Planning & Management
- Vol. 29
Predicting and measuring service disruption recovery time in railway gravity hump classification yards
- 14 February 2024
- Published by Elsevier in Journal of Rail Transport Planning & Management
- Vol. 29
Predicting and measuring service disruption recovery time in railway gravity hump classification yards
- Open Access
- 8 April 2023
- Published by SAGE Publications in Transportation Research Record: Journal of the Transportation Research Board
- Vol. 2677 (10) , 79-94
Quantifying the Influence of Volume Variability on Railway Hump Classification Yard Performance with AnyLogic Simulation
On North American freight railways, railcars transported in manifest (carload) freight trains are required to stop and be resorted at multiple intermediate classification (marshalling) yards before reaching their destination. Yard dwell time comprises the majority of total railcar transit time, and railway yard congestion can further promote mainline train delay. However, most of the literature on railroad performance has focused on the mainline, and few of the previous yard studies specifically examined the yard performance impacts of vast volume fluctuation because of variable shipping demand. To further understand yard behavior under these conditions, this paper explores the influence of inbound traffic volume variability on classification yard performance and outbound train size using a novel AnyLogic simulation model. A series of simulation experiments quantify the interaction between inbound and outbound train size variability as measured by different yard performance metrics under different overall traffic volumes and blocking patterns. The results quantify the impact of volume variation on railway classification yard level of service, and suggest that forming longer blocks leads to shorter dwell times but larger sensitivity to volume variation compared to forming multiple short blocks. Comparison of arrival and departure train size distributions indicates that classification yards moderate volume fluctuations while causing departure delays. These results serve as a quantitative foundation for further study of yard performance recovery patterns following traffic volume surges and the resulting interactions between yards and mainlines in the railway network, and can inform railroad operating plan adjustments to better manage yard traffic and utilize existing resources.
- 8 April 2023
- Published by SAGE Publications in Transportation Research Record: Journal of the Transportation Research Board
- Vol. 2677 (10) , 79-94
Open AccessQuantifying the Influence of Volume Variability on Railway Hump Classification Yard Performance with AnyLogic Simulation
On North American freight railways, railcars transported in manifest (carload) freight trains are required to stop and be resorted at multiple intermediate classification (marshalling) yards before reaching their destination. Yard dwell time comprises the majority of total railcar transit time, and railway yard congestion can further promote mainline train delay. However, most of the literature on railroad performance has focused on the mainline, and few of the previous yard studies specifically examined the yard performance impacts of vast volume fluctuation because of variable shipping demand. To further understand yard behavior under these conditions, this paper explores the influence of inbound traffic volume variability on classification yard performance and outbound train size using a novel AnyLogic simulation model. A series of simulation experiments quantify the interaction between inbound and outbound train size variability as measured by different yard performance metrics under different overall traffic volumes and blocking patterns. The results quantify the impact of volume variation on railway classification yard level of service, and suggest that forming longer blocks leads to shorter dwell times but larger sensitivity to volume variation compared to forming multiple short blocks. Comparison of arrival and departure train size distributions indicates that classification yards moderate volume fluctuations while causing departure delays. These results serve as a quantitative foundation for further study of yard performance recovery patterns following traffic volume surges and the resulting interactions between yards and mainlines in the railway network, and can inform railroad operating plan adjustments to better manage yard traffic and utilize existing resources.
- 21 December 2022
- Published by SAGE Publications in Transportation Research Record: Journal of the Transportation Research Board
- Vol. 2677 (5) , 793-811
Analysis of Derailment Severity Comparing Unit Trains at Transload Terminals and Manifest Trains at Railroad Switching and Hump Classification Yards
Growing demand for the safe and efficient transportation of crude oil and other flammable liquids by rail has highlighted the need to understand the relative derailment risk of two major heavy haul freight train types operating in the U.S.A.: unit trains dedicated to a single commodity between origin and destination terminals, and manifest trains transporting carload shipments between multiple classification (marshalling) yards. Despite the low likelihood of a train derailment, understanding the factors that influence the severity of a train derailment can potentially mitigate severe consequences and improve the reliability of railway transportation. Previous studies have quantified the derailment severity of mainline transportation but excluded derailments in yards and terminals. This paper examines U.S. freight train yard and terminal derailment data between 1996 and 2018 to develop a novel event-chain-based probabilistic risk analysis specifying yard and terminal arrival and departure derailment (derailments involving trains arriving/departing a yard or terminal) severity for both manifest trains and unit trains, and yard switching derailment (derailments during railcar sorting or classification in a yard) severity for manifest trains only. The approach developed in this paper provides predictions of severity with respect to the total number railcars derailed and hazardous materials (hazmat) railcars derailed given a manifest train or unit train shipment and hazmat railcar placement within the train consist. When combined with mainline derailment risk analysis methodologies and yard derailment rates from previous research, these yard and terminal derailment severity metrics facilitate more holistic comparisons of unit and manifest train derailment risks for all components of loaded trips.
- 21 December 2022
- Published by SAGE Publications in Transportation Research Record: Journal of the Transportation Research Board
- Vol. 2677 (5) , 793-811
Analysis of Derailment Severity Comparing Unit Trains at Transload Terminals and Manifest Trains at Railroad Switching and Hump Classification Yards
Growing demand for the safe and efficient transportation of crude oil and other flammable liquids by rail has highlighted the need to understand the relative derailment risk of two major heavy haul freight train types operating in the U.S.A.: unit trains dedicated to a single commodity between origin and destination terminals, and manifest trains transporting carload shipments between multiple classification (marshalling) yards. Despite the low likelihood of a train derailment, understanding the factors that influence the severity of a train derailment can potentially mitigate severe consequences and improve the reliability of railway transportation. Previous studies have quantified the derailment severity of mainline transportation but excluded derailments in yards and terminals. This paper examines U.S. freight train yard and terminal derailment data between 1996 and 2018 to develop a novel event-chain-based probabilistic risk analysis specifying yard and terminal arrival and departure derailment (derailments involving trains arriving/departing a yard or terminal) severity for both manifest trains and unit trains, and yard switching derailment (derailments during railcar sorting or classification in a yard) severity for manifest trains only. The approach developed in this paper provides predictions of severity with respect to the total number railcars derailed and hazardous materials (hazmat) railcars derailed given a manifest train or unit train shipment and hazmat railcar placement within the train consist. When combined with mainline derailment risk analysis methodologies and yard derailment rates from previous research, these yard and terminal derailment severity metrics facilitate more holistic comparisons of unit and manifest train derailment risks for all components of loaded trips.
- 15 June 2022
- Published by SAGE Publications in Transportation Research Record: Journal of the Transportation Research Board
- Vol. 2677 (1) , 311-325
Quantitative Derailment Rate Comparison of Unit Trains at Transload Terminals and Manifest Trains at Railroad Switching and Hump Classification Yards
The rapid expansion of demand for efficiently and safely transporting crude oil and other flammable liquids by rail in North America has highlighted the need to understand the relative derailment risk of two main freight train types operating in the United States (U.S.): unit trains and manifest trains. Previous studies have quantified the line-haul accident rates for these train types. However, the relative derailment likelihood of these two train types associated with train arrival/departure processes and yard switching operations has yet to be quantified. This study analyzes U.S. freight train traffic and yard/terminal derailment data between 1996 and 2018 by train type. For manifest trains, derailment rates are calculated per train arrival and departure in yards, and yard switching accidents per railcar handled in classification yards. For unit trains, the number of accidents per arrival and departure event in loading and unloading terminals is quantified. These rates are further refined to reflect particular unit train loading conditions and yard type for manifest trains. The analyses suggest that manifest trains have a four-times larger rate per yard arrival/departure than unit trains per terminal arrival/departure. Regardless of yard type, railcar shunting movements in classification yards are significant to the overall manifest train shipment derailment likelihood. An example case study demonstrates how a manifest train may be ten times more likely to derail than a unit train, while varying the number and type of intermediate yards for manifest train and the loading factor for unit train has distinct impacts on the overall derailment rate.
- 15 June 2022
- Published by SAGE Publications in Transportation Research Record: Journal of the Transportation Research Board
- Vol. 2677 (1) , 311-325
Quantitative Derailment Rate Comparison of Unit Trains at Transload Terminals and Manifest Trains at Railroad Switching and Hump Classification Yards
The rapid expansion of demand for efficiently and safely transporting crude oil and other flammable liquids by rail in North America has highlighted the need to understand the relative derailment risk of two main freight train types operating in the United States (U.S.): unit trains and manifest trains. Previous studies have quantified the line-haul accident rates for these train types. However, the relative derailment likelihood of these two train types associated with train arrival/departure processes and yard switching operations has yet to be quantified. This study analyzes U.S. freight train traffic and yard/terminal derailment data between 1996 and 2018 by train type. For manifest trains, derailment rates are calculated per train arrival and departure in yards, and yard switching accidents per railcar handled in classification yards. For unit trains, the number of accidents per arrival and departure event in loading and unloading terminals is quantified. These rates are further refined to reflect particular unit train loading conditions and yard type for manifest trains. The analyses suggest that manifest trains have a four-times larger rate per yard arrival/departure than unit trains per terminal arrival/departure. Regardless of yard type, railcar shunting movements in classification yards are significant to the overall manifest train shipment derailment likelihood. An example case study demonstrates how a manifest train may be ten times more likely to derail than a unit train, while varying the number and type of intermediate yards for manifest train and the loading factor for unit train has distinct impacts on the overall derailment rate.
- 15 September 2021
- Published by SAGE Publications in Transportation Research Record: Journal of the Transportation Research Board
- Vol. 2675 (12) , 890-903
Quantifying Recent Trends in Class 1 Freight Railroad Train Length and Weight by Train Type
Railroads have a strong economic incentive to maximize the length and weight of freight trains. Since the mid-1990s, various technological innovations have facilitated operation of longer and heavier trains, and railroads have made infrastructure investments to accommodate them. Recent shifts in railway operating and management strategies have placed added emphasis on long trains but have also drawn public and agency scrutiny. The advantages and disadvantages of increased train size are difficult to analyze because public data on train length and weight over time are limited. Articulated intermodal railcars, artificially short local trains, and light empty unit trains skew industry averages across all train types and mask trends over time. To provide greater insight on the average and distribution of train length and weight for different train types over time, the research team conducted a detailed analysis of Class 1 railroad annual report financial data and Surface Transportation Board waybill sample data collected for the years 1996 through 2018. Dividing traffic statistics by train type allows for a specific focus on loaded unit train length and weight distributions that isolates many factors skewing overall averages. Over the past 23 years, the average length and weight of loaded non-hazmat unit trains have steadily increased. Train size distributions indicate that unit trains exceeding 140 railcars in length have become more frequent over the past 10 years, whereas hazmat unit trains are typically smaller in size. This information can aid researchers and industry practitioners in assessing the benefits and disadvantages of operating longer trains.
- 15 September 2021
- Published by SAGE Publications in Transportation Research Record: Journal of the Transportation Research Board
- Vol. 2675 (12) , 890-903
Quantifying Recent Trends in Class 1 Freight Railroad Train Length and Weight by Train Type
Railroads have a strong economic incentive to maximize the length and weight of freight trains. Since the mid-1990s, various technological innovations have facilitated operation of longer and heavier trains, and railroads have made infrastructure investments to accommodate them. Recent shifts in railway operating and management strategies have placed added emphasis on long trains but have also drawn public and agency scrutiny. The advantages and disadvantages of increased train size are difficult to analyze because public data on train length and weight over time are limited. Articulated intermodal railcars, artificially short local trains, and light empty unit trains skew industry averages across all train types and mask trends over time. To provide greater insight on the average and distribution of train length and weight for different train types over time, the research team conducted a detailed analysis of Class 1 railroad annual report financial data and Surface Transportation Board waybill sample data collected for the years 1996 through 2018. Dividing traffic statistics by train type allows for a specific focus on loaded unit train length and weight distributions that isolates many factors skewing overall averages. Over the past 23 years, the average length and weight of loaded non-hazmat unit trains have steadily increased. Train size distributions indicate that unit trains exceeding 140 railcars in length have become more frequent over the past 10 years, whereas hazmat unit trains are typically smaller in size. This information can aid researchers and industry practitioners in assessing the benefits and disadvantages of operating longer trains.