Muhammad Usama
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
Research Subjects
Muhammad Usama has a total of 21 co-authors. Here are the top co-authors by number of shared publications:
- Haris N. Koutsopoulos
Department of Civil and Environmental Engineering, Northeastern University, Boston, MA
4 shared publications Onaira Zahoor
Southeast University,School of Transportation,Nanjing,China
4 shared publicationsYongjun Shen
Southeast University,School of Transportation,Nanjing,China
4 shared publicationsQiong Bao
Southeast University,School of Transportation,Nanjing,China
3 shared publications- Yongjun Shen
Southeast University
3 shared publications - Onaira Zahoor
School of Transportation, Southeast University, Nanjing 211189, China
2 shared publications
Latest Publications
- Open Access
- 27 January 2026
- Published by SAGE Publications in Transportation Research Record: Journal of the Transportation Research Board
- Vol. 2680 (10) , 568-585
Headway Management Strategies for High-Frequency Urban Rail Transit: Dynamic Approach to Real-Time Train Holding in Metro Systems
Urban rail transit systems frequently encounter challenges related to service reliability and passenger crowding, particularly during peak operational hours and in networks with complex service patterns. This paper presents an innovative approach to real-time train holding that addresses the unique challenges posed by systems with scheduled short-turning, where passenger loads at short-turning points can vary significantly. We developed a dual-strategy framework that combines (1) a real-time heuristic that calculates holding times using both historical data and real-time information to minimize passenger-experienced crowding, and (2) a predictive modeling approach that anticipates headway situations when full-length service trains from the terminal arrive at short-turning stations. Unlike conventional headway-equalizing strategies that overlook load variations in high-demand scenarios operating near capacity, our approach explicitly accounts for heterogeneous passenger loads across different service types to reduce denied boarding and passenger wait times. The effectiveness of our framework was evaluated using a microscopic simulation model of a high-frequency, high-demand urban rail transit system. The results demonstrate that the proposed approach reduced denied boarding incidents by 30% through improved train load balancing. The combination of predictive control with downstream holding strategies improved service quality through the proactive regulation of train dispatching at terminals, coupled with adjustments at key stations.
- 27 January 2026
- Published by SAGE Publications in Transportation Research Record: Journal of the Transportation Research Board
- Vol. 2680 (10) , 568-585
Open AccessHeadway Management Strategies for High-Frequency Urban Rail Transit: Dynamic Approach to Real-Time Train Holding in Metro Systems
Urban rail transit systems frequently encounter challenges related to service reliability and passenger crowding, particularly during peak operational hours and in networks with complex service patterns. This paper presents an innovative approach to real-time train holding that addresses the unique challenges posed by systems with scheduled short-turning, where passenger loads at short-turning points can vary significantly. We developed a dual-strategy framework that combines (1) a real-time heuristic that calculates holding times using both historical data and real-time information to minimize passenger-experienced crowding, and (2) a predictive modeling approach that anticipates headway situations when full-length service trains from the terminal arrive at short-turning stations. Unlike conventional headway-equalizing strategies that overlook load variations in high-demand scenarios operating near capacity, our approach explicitly accounts for heterogeneous passenger loads across different service types to reduce denied boarding and passenger wait times. The effectiveness of our framework was evaluated using a microscopic simulation model of a high-frequency, high-demand urban rail transit system. The results demonstrate that the proposed approach reduced denied boarding incidents by 30% through improved train load balancing. The combination of predictive control with downstream holding strategies improved service quality through the proactive regulation of train dispatching at terminals, coupled with adjustments at key stations.
- Open Access
Estimating link level traffic emissions: enhancing MOVES with open-source data
Open-source data offers a scalable and transparent foundation for estimating vehicle activity and emissions in urban regions. In this study, we propose a data-driven framework that integrates MOVES and open-source GPS trajectory data, OpenStreetMap (OSM) road networks, regional traffic datasets and satellite imagery-derived feature vectors to estimate the link level operating mode distribution and traffic emissions. A neural network model is trained to predict the distribution of MOVES-defined operating modes using only features derived from readily available data. The proposed methodology was applied using open-source data related to 45 municipalities in the Boston Metropolitan area. The "ground truth" operating mode distribution was established using OSM open-source GPS trajectories. Compared to the MOVES baseline, the proposed model reduces RMSE by over 50% for regional scale traffic emissions of key pollutants including CO, NOx, CO2, and PM2.5. This study demonstrates the feasibility of low-cost, replicable, and data-driven emissions estimation using fully open data sources.
Open AccessEstimating link level traffic emissions: enhancing MOVES with open-source data
Open-source data offers a scalable and transparent foundation for estimating vehicle activity and emissions in urban regions. In this study, we propose a data-driven framework that integrates MOVES and open-source GPS trajectory data, OpenStreetMap (OSM) road networks, regional traffic datasets and satellite imagery-derived feature vectors to estimate the link level operating mode distribution and traffic emissions. A neural network model is trained to predict the distribution of MOVES-defined operating modes using only features derived from readily available data. The proposed methodology was applied using open-source data related to 45 municipalities in the Boston Metropolitan area. The "ground truth" operating mode distribution was established using OSM open-source GPS trajectories. Compared to the MOVES baseline, the proposed model reduces RMSE by over 50% for regional scale traffic emissions of key pollutants including CO, NOx, CO2, and PM2.5. This study demonstrates the feasibility of low-cost, replicable, and data-driven emissions estimation using fully open data sources.
- Open Access
Real Time Headway Predictions in Urban Rail Systems and Implications for Service Control: A Deep Learning Approach
Efficient real-time dispatching in urban metro systems is essential for ensuring service reliability, maximizing resource utilization, and improving passenger satisfaction. This study presents a novel deep learning framework centered on a Convolutional Long Short-Term Memory (ConvLSTM) model designed to predict the complex spatiotemporal propagation of train headways across an entire metro line. By directly incorporating planned terminal headways as a critical input alongside historical headway data, the proposed model accurately forecasts future headway dynamics, effectively capturing both their temporal evolution and spatial dependencies across all stations. This capability empowers dispatchers to evaluate the impact of various terminal headway control decisions without resorting to computationally intensive simulations. We introduce a flexible methodology to simulate diverse dispatcher strategies, ranging from maintaining even headways to implementing custom patterns derived from observed terminal departures. In contrast to existing research primarily focused on passenger load predictioning or atypical disruption scenarios, our approach emphasizes proactive operational control. Evaluated on a large-scale dataset from an urban metro line, the proposed ConvLSTM model demonstrates promising headway predictions, offering actionable insights for real-time decision-making. This framework provides rail operators with a powerful, computationally efficient tool to optimize dispatching strategies, thereby significantly improving service consistency and passenger satisfaction.
Open AccessReal Time Headway Predictions in Urban Rail Systems and Implications for Service Control: A Deep Learning Approach
Efficient real-time dispatching in urban metro systems is essential for ensuring service reliability, maximizing resource utilization, and improving passenger satisfaction. This study presents a novel deep learning framework centered on a Convolutional Long Short-Term Memory (ConvLSTM) model designed to predict the complex spatiotemporal propagation of train headways across an entire metro line. By directly incorporating planned terminal headways as a critical input alongside historical headway data, the proposed model accurately forecasts future headway dynamics, effectively capturing both their temporal evolution and spatial dependencies across all stations. This capability empowers dispatchers to evaluate the impact of various terminal headway control decisions without resorting to computationally intensive simulations. We introduce a flexible methodology to simulate diverse dispatcher strategies, ranging from maintaining even headways to implementing custom patterns derived from observed terminal departures. In contrast to existing research primarily focused on passenger load predictioning or atypical disruption scenarios, our approach emphasizes proactive operational control. Evaluated on a large-scale dataset from an urban metro line, the proposed ConvLSTM model demonstrates promising headway predictions, offering actionable insights for real-time decision-making. This framework provides rail operators with a powerful, computationally efficient tool to optimize dispatching strategies, thereby significantly improving service consistency and passenger satisfaction.
- Open Access
- 20 August 2025
- Published by SAGE Publications in Transportation Research Record: Journal of the Transportation Research Board
- Vol. 2679 (11) , 511-525
Estimating City-wide Operating Mode Distribution of Light-Duty Vehicles: A Neural Network-based Approach
Driving cycles are a set of driving conditions and are crucial for the existing emission estimation model to evaluate vehicle performance, fuel efficiency, and emissions, by matching them with average speed to calculate the operating modes, such as braking, idling, and cruising. Although existing emission estimation models, such as the Motor Vehicle Emission Simulator (MOVES), are powerful tools, their reliance on predefined driving cycles can be limiting, as these cycles often do not accurately represent regional driving conditions, making the models less effective for city-wide analyses. To solve this problem, this paper proposes a modular neural network-based framework to estimate operating mode distributions bypassing the driving cycle development phase, utilizing macroscopic variables such as speed, flow, and link infrastructure attributes. The proposed method is validated using a well-calibrated microsimulation model of Brookline MA, the United States. The results indicate that the proposed framework outperforms the operating mode distribution calculated by MOVES based on default driving cycles, providing a closer match to the actual operating mode distribution derived from trajectory data obtained from the microsimulation. Specifically, the proposed model achieves an average RMSE of 0.04 in predicting operating mode distribution, compared with 0.08 for MOVES. The average error in emission estimation across pollutants is 8.57% for the proposed method, lower than the 32.86% error for MOVES. In particular, for the estimation of CO 2 , the proposed method has an error of just 4%, compared with 35% for MOVES. The proposed model can be utilized for real-time emissions monitoring by providing rapid and accurate emissions estimates with easily accessible inputs.
- 20 August 2025
- Published by SAGE Publications in Transportation Research Record: Journal of the Transportation Research Board
- Vol. 2679 (11) , 511-525
Open AccessEstimating City-wide Operating Mode Distribution of Light-Duty Vehicles: A Neural Network-based Approach
Driving cycles are a set of driving conditions and are crucial for the existing emission estimation model to evaluate vehicle performance, fuel efficiency, and emissions, by matching them with average speed to calculate the operating modes, such as braking, idling, and cruising. Although existing emission estimation models, such as the Motor Vehicle Emission Simulator (MOVES), are powerful tools, their reliance on predefined driving cycles can be limiting, as these cycles often do not accurately represent regional driving conditions, making the models less effective for city-wide analyses. To solve this problem, this paper proposes a modular neural network-based framework to estimate operating mode distributions bypassing the driving cycle development phase, utilizing macroscopic variables such as speed, flow, and link infrastructure attributes. The proposed method is validated using a well-calibrated microsimulation model of Brookline MA, the United States. The results indicate that the proposed framework outperforms the operating mode distribution calculated by MOVES based on default driving cycles, providing a closer match to the actual operating mode distribution derived from trajectory data obtained from the microsimulation. Specifically, the proposed model achieves an average RMSE of 0.04 in predicting operating mode distribution, compared with 0.08 for MOVES. The average error in emission estimation across pollutants is 8.57% for the proposed method, lower than the 32.86% error for MOVES. In particular, for the estimation of CO 2 , the proposed method has an error of just 4%, compared with 35% for MOVES. The proposed model can be utilized for real-time emissions monitoring by providing rapid and accurate emissions estimates with easily accessible inputs.
- Open Access
- 27 November 2022
- Published by MDPI AG in Algorithms
- Vol. 15 (12) , 447
Physics-Informed Neural Networks (PINNs)-Based Traffic State Estimation: An Application to Traffic Network
Traffic state estimation (TSE) is a critical component of the efficient intelligent transportation systems (ITS) operations. In the literature, TSE methods are divided into model-driven methods and data-driven methods. Each approach has its limitations. The physics information-based neural network (PINN) framework emerges to mitigate the limitations of the traditional TSE methods, while the state-of-art of such a framework has focused on single road segments but can hardly deal with traffic networks. This paper introduces a PINN framework that can effectively make use of a small amount of observational speed data to obtain high-quality TSEs for a traffic network. Both model-driven and data-driven components are incorporated into PINNs to combine the advantages of both approaches and to overcome their disadvantages. Simulation data of simple traffic networks are used for studying the highway network TSE. This paper demonstrates how to solve the popular LWR physical traffic flow model with a PINN for a traffic network. Experimental results confirm that the proposed approach is promising for estimating network traffic accurately.
- 27 November 2022
- Published by MDPI AG in Algorithms
- Vol. 15 (12) , 447
Open AccessPhysics-Informed Neural Networks (PINNs)-Based Traffic State Estimation: An Application to Traffic Network
Traffic state estimation (TSE) is a critical component of the efficient intelligent transportation systems (ITS) operations. In the literature, TSE methods are divided into model-driven methods and data-driven methods. Each approach has its limitations. The physics information-based neural network (PINN) framework emerges to mitigate the limitations of the traditional TSE methods, while the state-of-art of such a framework has focused on single road segments but can hardly deal with traffic networks. This paper introduces a PINN framework that can effectively make use of a small amount of observational speed data to obtain high-quality TSEs for a traffic network. Both model-driven and data-driven components are incorporated into PINNs to combine the advantages of both approaches and to overcome their disadvantages. Simulation data of simple traffic networks are used for studying the highway network TSE. This paper demonstrates how to solve the popular LWR physical traffic flow model with a PINN for a traffic network. Experimental results confirm that the proposed approach is promising for estimating network traffic accurately.
- Open Access
- 29 November 2020
- Published by MDPI AG in International Journal of Environmental Research and Public Health
- Vol. 17 (23) , 8886
Assessing Fitness-To-Drive among Older Drivers: A Comparative Analysis of Potential Alternatives to on-Road Driving Test
To enable older drivers to maintain mobility without endangering public safety, it is necessary to develop more effective means of assessing their fitness-to-drive as alternatives to an on-road driving test. In this study, a functional ability test, simulated driving test, and on-road driving test were carried out for 136 older drivers. Influencing factors related to fitness-to-drive were selected based on the correlation between the outcome measure of each test and the pass/fail outcome of the on-road driving test. Four potential alternatives combining different tests were considered and three modeling techniques were compared when constructing the fitness-to-drive assessment model for the elderly. As a result, 92 participants completed all of the tests, of which 61 passed the on-road driving test and the remaining 31 failed. A total of seven influencing factors from all types of tests were selected. The best model was trained by the technique of gradient boosted machine using all of the seven factors, generating the highest accuracy of 92.8%, with sensitivity of 0.94 and specificity of 0.90. The proposed fitness-to-drive assessment method is considered an effective alternative to the on-road driving test, and the results offer a valuable reference for those unfit-to-drive older drivers to either adjust their driving behavior or cease driving.
- 29 November 2020
- Published by MDPI AG in International Journal of Environmental Research and Public Health
- Vol. 17 (23) , 8886
Open AccessAssessing Fitness-To-Drive among Older Drivers: A Comparative Analysis of Potential Alternatives to on-Road Driving Test
To enable older drivers to maintain mobility without endangering public safety, it is necessary to develop more effective means of assessing their fitness-to-drive as alternatives to an on-road driving test. In this study, a functional ability test, simulated driving test, and on-road driving test were carried out for 136 older drivers. Influencing factors related to fitness-to-drive were selected based on the correlation between the outcome measure of each test and the pass/fail outcome of the on-road driving test. Four potential alternatives combining different tests were considered and three modeling techniques were compared when constructing the fitness-to-drive assessment model for the elderly. As a result, 92 participants completed all of the tests, of which 61 passed the on-road driving test and the remaining 31 failed. A total of seven influencing factors from all types of tests were selected. The best model was trained by the technique of gradient boosted machine using all of the seven factors, generating the highest accuracy of 92.8%, with sensitivity of 0.94 and specificity of 0.90. The proposed fitness-to-drive assessment method is considered an effective alternative to the on-road driving test, and the results offer a valuable reference for those unfit-to-drive older drivers to either adjust their driving behavior or cease driving.
- Open Access
- 28 January 2020
- Published by Center for Transportation Studies in Journal of Transport and Land Use
- Vol. 13 (1) , 491-515
Dockless bike-sharing system: Solving the problem of faulty bikes with simultaneous rebalancing operation
The Free-Floating Bike-Sharing System (FFBS) enables commuters to pick up and drop off a shared bike without going to a docking station before and after the trip. However, the specific sharing features of the FFBS can also be a problem for users because randomly scattered and faulty bikes within the system can result in bike unavailability and maintenance difficulties. This not only negatively impacts the company’s service quality but also causes user safety issues. In this study, a mechanism for the rebalancing of useable bikes and faulty bikes is presented in two steps: 1) gathering each faulty bike at a station determined by the model traversing the shortest path by light service vehicles; and 2) enabling the rebalancing operation to obtain optimal bike inventory levels at all stations and collect faulty bikes at a depot. The destination station from which each faulty bike is taken is considered a decision variable rather than shifting them to a closer station. The mechanism is based on minimizing the total cost of the rebalancing operation by optimizing the formulated problem, supported by time and capacity constraints. Using CPLEX solver, the validity of the model is tested through different numerical experiments.
- 28 January 2020
- Published by Center for Transportation Studies in Journal of Transport and Land Use
- Vol. 13 (1) , 491-515
Open AccessDockless bike-sharing system: Solving the problem of faulty bikes with simultaneous rebalancing operation
The Free-Floating Bike-Sharing System (FFBS) enables commuters to pick up and drop off a shared bike without going to a docking station before and after the trip. However, the specific sharing features of the FFBS can also be a problem for users because randomly scattered and faulty bikes within the system can result in bike unavailability and maintenance difficulties. This not only negatively impacts the company’s service quality but also causes user safety issues. In this study, a mechanism for the rebalancing of useable bikes and faulty bikes is presented in two steps: 1) gathering each faulty bike at a station determined by the model traversing the shortest path by light service vehicles; and 2) enabling the rebalancing operation to obtain optimal bike inventory levels at all stations and collect faulty bikes at a depot. The destination station from which each faulty bike is taken is considered a decision variable rather than shifting them to a closer station. The mechanism is based on minimizing the total cost of the rebalancing operation by optimizing the formulated problem, supported by time and capacity constraints. Using CPLEX solver, the validity of the model is tested through different numerical experiments.
- 1 November 2019
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 648-653
Assessing Fitness to Drive among Older Adults using Random Forest
To meet the requirements of assessing fitness-to-drive among older adults, recently several studies have validated the use of driving simulator. This study adheres to the opinion of developing a method to assess fitness to drive based on various assessments and driving simulator rides using Random Forest. The goals are: a) to evaluate older drivers’ fïtness-to-drive based on different assessments, b) to determine which type of assessment, or combination of assessment provides the best prediction of the fïtness-to-drive, c) to determine the validity of driving simulator performance in predicting fitness to drive. The criterion of fitness to drive is determined in an on-road driving assessment by specialized evaluator from Center for Determination of Fitness to Drive and Car Adaptations (CARA). The best model acquires an accuracy of 82.2% on average, making the approach of combining all predictors from different assessments acceptable and valuable for assessing fitness-to-drive among older drivers.
- 1 November 2019
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 648-653
Assessing Fitness to Drive among Older Adults using Random Forest
To meet the requirements of assessing fitness-to-drive among older adults, recently several studies have validated the use of driving simulator. This study adheres to the opinion of developing a method to assess fitness to drive based on various assessments and driving simulator rides using Random Forest. The goals are: a) to evaluate older drivers’ fïtness-to-drive based on different assessments, b) to determine which type of assessment, or combination of assessment provides the best prediction of the fïtness-to-drive, c) to determine the validity of driving simulator performance in predicting fitness to drive. The criterion of fitness to drive is determined in an on-road driving assessment by specialized evaluator from Center for Determination of Fitness to Drive and Car Adaptations (CARA). The best model acquires an accuracy of 82.2% on average, making the approach of combining all predictors from different assessments acceptable and valuable for assessing fitness-to-drive among older drivers.
- 2 July 2019
- Published by American Society of Civil Engineers (ASCE)
- p. 4963-4974
Dockless Bike-Sharing Rebalancing Problem with Simultaneous Faulty Bike Recycling
- 2 July 2019
- Published by American Society of Civil Engineers (ASCE)
- p. 4963-4974
Dockless Bike-Sharing Rebalancing Problem with Simultaneous Faulty Bike Recycling
- 2 July 2019
- Published by American Society of Civil Engineers (ASCE)
- p. 4156-4168
LNG Bus Emissions Prediction Using Neural Network
Being a key feature in any developed society, transportation systems play a vital role to satisfy mobility and accessibility needs. Urbanization, an expanding transport sector and timely access to any specific location are the pivotal factors that contribute to the negative impact on environment. Vehicular emissions such as carbon monoxide (CO), nitrogen oxides (NOX), hydrocarbons (HC), and carbon dioxide (CO2) are anthropogenic in nature. This study corroborates the use of a neural network model for the prediction of vehicular emissions based on actual road data of LNG buses measured by portable emissions measurement system, running on line no. 51 in Zhenjiang, China. The data encompasses speed, acceleration, road grade, and passenger load as the inputs and indicative emissions (CO2, NOx, CO, and HC) as outputs. This research aims at LNG transit bus emissions modeling based on vehicular and road parameters to measure vehicle specific power correlation with vehicular emissions.
- 2 July 2019
- Published by American Society of Civil Engineers (ASCE)
- p. 4156-4168
LNG Bus Emissions Prediction Using Neural Network
Being a key feature in any developed society, transportation systems play a vital role to satisfy mobility and accessibility needs. Urbanization, an expanding transport sector and timely access to any specific location are the pivotal factors that contribute to the negative impact on environment. Vehicular emissions such as carbon monoxide (CO), nitrogen oxides (NOX), hydrocarbons (HC), and carbon dioxide (CO2) are anthropogenic in nature. This study corroborates the use of a neural network model for the prediction of vehicular emissions based on actual road data of LNG buses measured by portable emissions measurement system, running on line no. 51 in Zhenjiang, China. The data encompasses speed, acceleration, road grade, and passenger load as the inputs and indicative emissions (CO2, NOx, CO, and HC) as outputs. This research aims at LNG transit bus emissions modeling based on vehicular and road parameters to measure vehicle specific power correlation with vehicular emissions.