Presentations

Recording

Cal Skiles- Impact of Delivery Vehicles on Roadways in Residential Areas

Abstract: Communications technologies and e-commerce have profound effects on travel behavior, including shopping trips. Such changes are, in turn, transforming operations associated with delivery of goods and services, and consequently, on the use of transportation infrastructure, including residential roads. As e-commerce continues to grow, including the spike from pandemic-era lockdowns, an effect of the growth of home deliveries is the potential impact on pavement performance from the increased number and weight of delivery vehicles. Delivery vehicles are heavier than personal vehicles, and thus, they have an outsized impact on pavement wear, especially in the case of residential streets designed for limited traffic volumes. A method of accounting for the increased heavier traffic is presented as slightly changing the pavement design such that the maintenance schedule of these roads remains the same. A doubling of the number of delivery vehicles leads to a shorter useful life of the pavement, and then a corresponding increase in the surface thickness of the pavement allows for the maintenance schedule to remain unchanged was determined to be an increase of 6.35mm (0.25″) in the pavement surface thickness. Each 6.35mm (0.25″) increase in pavement surface thickness has a material cost increase of $10,874 per km ($17,500 per mile) of roadway for an average residential street. The analysis is then repeated for the increased weight of electric delivery vehicles, and an even greater pavement surface thickness is needed to keep the maintenance frequency consistent.

Bio: Cal Skiles is a third-year Ph.D. candidate at Northwestern University. He completed is undergraduate studies at the University of Illinois Urbana-Champaign. His current research is focused on infrastructure management, planning, and operations.

Recording

Amir Fard- Hierarchical Deep Reinforcement Learning Framework for Optimizing Cross-Asset Budget Allocation in Municipal Asset Management

Abstract: Efficiently managing municipal budgets requires innovative strategies for allocating funds across a broad spectrum of asset classes, including sewer systems and pavement networks, and their respective subclasses, such as Arterial and Collector roads, which often derive from varied funding sources. This study introduces an innovative method for optimizing cross-asset budget allocation through a hierarchical deep reinforcement learning framework. Utilizing the soft-actor critic algorithm at the system level, our approach facilitates dynamic and adaptive distribution of budgets across and within asset classes, while linear programming at the asset level ensures optimal fund utilization. This combined strategy allows for customized budget allocation, catering to the specific needs and priorities of different asset classes or subclasses, regardless of their functional similarities or disparate funding sources. Our methodology demonstrates marked improvements in budget allocation efficiency and asset performance, highlighting the potential for refined and effective MRR planning and financial management within municipal infrastructure.

Bio: Amir Fard is a PhD candidate in the Civil Engineering Department at Toronto Metropolitan University, Canada. He holds bachelor’s and master’s degrees in Aerospace Engineering from Sharif University of Technology.  Under the supervision of Dr. Arnold Yuan, his research focuses on Risk-informed infrastructure life-cycle planning. Currently, he is working on developing a multi-agent deep reinforcement learning framework for adaptive inspection and maintenance plan optimization for infrastructure networks.

Rahul- Quality assessment of pavement condition monitoring
data of an Indian highway

Abstract: The use of network survey vehicles (NSVs) has accelerated in recent years for the functional evaluation of Highway Assets. However, due to the lack of standardization in condition monitoring process, there is inherent variability in documented data, over space and time, due to the usage of multiple equipment, varying segment lengths, etc. This study aims to assess the data consistency of International Roughness Index (IRI) values across different NSV operators. In order to evaluate the reproducibility and repeatability of these devices, fifteen runs were conducted using three Network Survey Vehicles along a 6 km section of National Highway 44 (NH44). The repeatability of NSV devices was evaluated using coefficient of variation (CoV) associated with the IRI estimates for different segment lengths and number of NSVs runs being averaged to determine segment IRI. The results indicate that CoV reduces as the number of runs and the segment length under consideration increases. The results across NSV devices suggests differences in sensitivity and measurement errors. The run-to-run variability is also observed to be a function of the average IRI of the segment. Finally, some exploratory analysis to uncover the sources of variation are discussed.

Bio: 2nd year M.Tech. transportation engineering student from Indian Institute of Technology, Kanpur. Completed B.E. in Civil engineering from Chandigarh University, Punjab in 2021.

Thesis supervisor: Dr. Aditya Medury.

Research Interests: Machine learning (prediction models) and data analytics.

Recording

Babatunde Atolagbe – A two-step exploratory framework for enhancing the explicability and informing efficient use of network-based decision support tools for transportation asset management

Abstract: Decision support models are widely used by asset managers to make effective and efficient decisions related to the maintenance, repair, and renewal of aging transportation infrastructure. These models range from simple rule-based tools, such as decision trees in pavement management systems, to complex blackbox models based on sophisticated optimization frameworks. While blackbox models can provide near optimal solutions, the analytical burden and lack of clarity of the modeling process often discourage asset managers from using them. Several inputs and parameters are commonly introduced in the models to achieve realism. Unfortunately, these variables influence the outcomes of the models in ways that may not always be as anticipated. Also, the enormous computational cost that results from attempting to adapt the modeling process to real life transportation networks usually forces model developers to use toy networks for illustration thereby questioning the realism of the models. This research explores a solution to this bipartite challenge. First, the relative importance of inputs and/or parameters in a complex model is explored using a variance-based global sensitivity analysis technique, the Sobol technique. A case-study model for determining the optimal lifecycle costs of maintenance of roadways considering traffic disruption costs is used for illustration. Preliminary results show that complex models can be simplified by eliminating or setting the least important parameters to pre-determined values. Second, a model selection framework referred to as a “model-of-models” is proposed. The model-of-models considers three models of varying complexities in the same class of infrastructure problem. Decision scenarios, reflecting
random combinations of the inputs to the support tool, global structural indices for random
transportation networks as well as varying levels of traffic demand in these networks are
generated. Each of these scenarios are then assigned unique decision support tool based on the efficiency of the tool for the specific scenario. With the scenarios labeled, a classification algorithm is trained to learn the mapping between the scenarios and the assigned tools thus resulting in an upper-level model that can be used to determine the appropriate tool for new scenarios.

Bio: Babatunde Atolagbe is a PhD student in the Department of Civil and Environmental Engineering at the University of Delaware. His dissertation focuses on using machine learning to identify the appropriate decision support tools for asset management to different problem contexts. He completed a bachelor’s degree from the University of Ilorin, Nigeria, and a master’s degree from University of Tennessee, Chattanooga, both in Civil Engineering.

He has worked on a variety of research projects including the development of cityscape mesh generation program (algorithm) for environmental fluids dynamics problems, strategic compensation of toll road operations to relieve maintenance impacts, and prioritization and planning of multi-modal transportation facilities. He also has work experience with the Maryland Department of Transportation and is a licensed Professional Engineer in Maryland.

Dadi Bhargavi –  Comparison of survival analysis and logistic regression methods for pavement crack initiation modeling

Abstract: Road infrastructure plays an important role in economic development, to ensure the  safety and effective preservation of road infrastructure periodic maintenance and rehabilitation is  necessary to prevent irreversible deterioration. To take the appropriate action at the correct time  there is a need for pavement performance prediction. But the pavement inspection is censored  which implies that we cannot observe the crack initiation time for all sections in the study, to  account for this incomplete information survival analysis is used which is commonly used to model  time to-event data, however survival analysis techniques fail to capture the nonlinear interactions  between the dependent variable and covariates. Survival stacking, which was less explored in  pavement infrastructure, provides an alternate approach to model the survival analysis problem as  a classification problem by reshaping the survival dataset. This allows the use of already existing  classification algorithms for right censored data. We have performed survival stacking with logistic  regression and compared the results with cox model on empirical and simulated data.

Bio: Graduated with a bachelor’s degree in civil engineering from NIT Trichy. Currently  pursuing a master’s degree in Transportation Engineering at IIT Kanpur under the guidance of Dr.  Aditya Medury, with a focus on pavement engineering and predictive modeling. Current research  includes comparison of survival analysis and logistic regression models for pavement crack  initiation prediction. 

Yuto Nakazato-Optimal annual work-zone schedule by days and sections for road networks considering work-zone synchronization and user cost

Abstract: In asset management of road networks, intervention for the facilities is determined based on a long-term management plan. To carry out all interventions each day, it is desirable to appropriately consider the economies of scale in intervention costs and user costs arising from traffic regulations for intervention.This study formulates an optimal daily work-zone schedule at a 100-meter sectional level for road networks, with the goal of minimizing the sum of user costs and intervention costs. To alleviate computational load, this study introduces a bi-level optimization method that divides the problem into a lower-level problem at the 100-meter sectional level within links, and an upper-level problem at the link level within the entire road network. Numerical case studies demonstrate the proposed approach’s capability to derive an optimal work-zone schedule for a real-scale network.

Bio: Yuto Nakazato is a second-year Ph.D. student at Tohoku University in Miyagi, Japan, where he also completed his Bachelor’s and Master’s degrees. Since his undergraduate days, he has been involved in mathematical optimization for infrastructure intervention policy. His current research focuses on the development of intervention policies for large-scale infrastructure systems, particularly in the context of road networks.

Juan del Campo-Sustainable Paved Infrastructure Management: Optimization of Global Warming Impact and Condition Performance with Dynamic Programming

Abstract: This work proposes a sustainable infrastructure management tool to support transport agencies in achieving two performance goals for roadway networks: maximizing improvements in the physical condition of their infrastructure and minimizing their global warming impact. The proposed tool accounts for a range of uncertainties inherent in the management of roadway assets and is able to identify flexible management strategies that can adapt with future conditions. The case study, which is based on real roadway network data, demonstrates the efficacy of the proposed tool in identifying sustainable management strategies and the effect of policy choices on the overall performance of a roadway system.

Bio: Juan del Campo is a first-year Ph.D. student at the University of British Columbia (UBC) in Vancouver, Canada. He earned his bachelor’s degree in his hometown (Puebla, Mexico), and then his master’s degree at UBC. Despite still struggling with the colder temperatures, Juan has found his passion in research in Vancouver and UBC. His research focus for his Ph.D. is the interface of sustainable asset management with machine learning.

Arnór B. Elvarsson-Responsive infrastructure planning organizations to meet long-term societal needs

Abstract: Planning in space requires consideration of land, mobility and infrastructure. For example, when a land parcel is to be developed for housing residents, those residents will have daily mobility demands and consequently, road and rail infrastructure are required to serve those demands in alignment with the spatial development. If planners decide to develop a particular parcel, a long time can pass – e.g., during public deliberations, permit applications and land acquisition – until the necessary infrastructure is constructed. The process can take even longer when infrastructure is planned on a regional level: if regions expect high population growth, planners may observe that the available mobility services will not be able to provide the necessary infrastructure capacity (e.g., frequent enough train service to carry peak hour demand, or large enough highways). The large number of stakeholders involved and the shaping impact that the infrastructure will have on the region over a long-term uncertain planning horizon, can slow down the process of expanding rail service or widening a highway. Meanwhile, mobility users may feel discomfort due to being stuck in traffic or in overcrowded transit services for longer times than anticipated. It might benefit users and other stakeholders if planning organisations were able to meet changing stakeholder needs more quickly.
This talk will (1) address the need to define responsiveness of planning organisations, (2) report on results for a case study for responsive planning organisations and (3) discuss the implications this has on research for infrastructure planning and related decision-support tools.

Bio: Arnór is a civil engineer experienced in the planning of transportation infrastructure. He holds an MSc in Spatial Development and Infrastructure Systems from ETH Zürich after a degree in Civil and Environmental Engineering from the University of Iceland. After some experience in practice, Arnór returned to ETH Zürich to research the planning of infrastructure systems. His research is part of a large research module on the interconnected planning of mobility, infrastructure and land-use at Future Cities Lab – Global in Singapore. Arnór’s contribution is based on the thesis that infrastructure planning processes can be made shorter in duration. This is challenging because there is a large number of stakeholders involved and the infrastructure will have a shaping impact on the region over a long-term planning horizon that is largely uncertain. Motivated by global changes that may largely impact societal needs, Arnór’s doctorate is about addressing the time duration of infrastructure planning, mapping the planning process, providing decision-support for modelling effective planning outcomes and how to select and rank planning decisions considering both effectiveness and efficiency of planning.

Priyadarshan Patil-Medium- and Heavy-duty vehicle electrification: current planning challenges and insights

Abstract: The strong legislative push for medium- and heavy-duty vehicle electrification is leading to strategic and operational planning within the ecosystem. This talk explores some of the challenges faced by fleet owners, utility companies, and third party resource providers during this planning process. Specifically, we talk about the vehicle and charger sizing problems and operational planning issues for fleet owners, demand identification and forecasting for utility companies, and lastly, travel demand modeling for third-party high-speed charging providers. This work helps provide insights into some of the challenges facing these agencies and some approaches to tackle these issues.

Bio: Priyadarshan Patil is an Operations Research Scientist at Electrotempo Inc working on fleet electrification. He has a Ph.D. in Operations Research from The University of Texas at Austin, focusing on traffic assignment models and solution algorithms. His research focuses on network optimization, electrification, clean energy, and resilience.

Aaron Chow-Coastal Protection Strategies to Minimize Transportation Network Disruption from Sea Level Rise in Abu Dhabi

Abstract: As sea levels rise, there has been an increase in research focused on the protection of shoreline infrastructure and transportation systems that may increasingly suffer from permanent capacity and accessibility reduction. This talk focuses on Abu Dhabi, UAE, a city vulnerable to inundation due to its network of low-lying islands and identifies protection strategies that will minimize transportation network delays. The model considers hydrodynamic interactions and traffic assignment. The results show some shoreline portions are critical, and their protection leads to lower congestion, while there are combinations of shoreline protection that worsen the congestion levels. The results also show that in some cases, the marginal effects of protecting one precinct may yield a better reduction of congestion than multiple other precincts. This research can provide a general framework for the protection of transportation infrastructure against sea level rise.

Bio: Aaron Chow is a Research Scientist with the Transportation Infrastructure Management Lab, Engineering Division, at NYU Abu Dhabi, UAE. Previously he worked as a postdoctoral associate at MIT Civil and Environmental Engineering, focusing on the transport and dispersion of excess salinity in coastal discharges from desalination plants in Kuwait. Before this he worked as an environmental consultant at Gradient and Berkeley Research Group in Boston, MA.

Meet Saiya-Incorporating economies of scale in top-down pavement management system

Abstract: Maintenance, rehabilitation and reconstruction (MR&R) decision-making in the context of infrastructure asset management is a multi-year planning process that seeks to incorporate agency and user considerations over a planning horizon. Markov decision process (MDP) is a prominent class of modeling framework that is extensively employed in the top-down MR&R planning literature to solve this resource allocation problem using randomized policies. Randomized policies are typically modeled within the top-down approaches by assuming that the agency incurs unit costs for any maintenance action. However, there is empirical evidence to suggest that economies-of-scale (EoS) are present when agencies implement MR&R actions. Economies of scale refers to reduction in unit costs as the quantity of actions of similar nature increases. This work introduces EoS within a system-level, top-down MR&R optimization problem. In particular, given the concave nature of the objective function owing to the introduction of EoS, two solution frameworks are implemented: a piecewise linear (PWL) approximation technique and a sigmoidal programming approach. Various attributes and solution methods for the problem formulation are discussed and comparisons between unit cost and proposed models are made. Comparisons are also made between different solution methods and their efficiencies. Changes in decision making because of incorporating economies of scale are then highlighted and discussed in the results.

Bio: Meet completed his MTech from Indian Institute of Technology, Kanpur specialising in Infrastructure Engineering and Management with institute recognition for academic excellence. His thesis focused on incorporating economies of scale in pavement asset management framework. During his internship at Bureau of Indian Standards, GoI he examined various ISO, ASTM, Eurocode standards along with latest literature to make recommendations and help improve 30+ indigenous BIS standards to incorporate sustainability into codal provisions. Post his MTech, he joined AECOM – Enterprise capabilities centre as a junior bridge engineer in Rails structure team.

Max Ng-Redesigning Large-Scale Multimodal Transit Networks with Shared Autonomous Mobility Services

Abstract: Public transit systems have faced challenges and opportunities from emerging Shared Autonomous Mobility Services (SAMS). This study addresses a city-scale multimodal transit network design problem, with shared autonomous vehicles as both transit feeders and a direct interzonal mode. The framework captures spatial demand and modal characteristics, considers intermodal transfers and express services, determines transit infrastructure investment and path flows, and designs transit routes. A system-optimal multimodal transit network is designed with minimum total door-to-door generalized costs of users and operators, while satisfying existing transit origin-destination demand within a pre-set infrastructure budget. Firstly, the geography, demand, and modes in each clustered zone are characterized with continuous approximation. Afterward, the decisions of network link investment and multimodal path flows in zonal connection optimization are formulated as a minimum-cost multi-commodity network flow (MCNF) problem and solved efficiently with a mixed-integer linear programming (MILP) solver. Subsequently, the route generation problem is solved by expanding the MCNF formulation to minimize intramodal transfers. To demonstrate the framework efficiency, this study uses transit demand from the Chicago metropolitan area to redesign a multimodal transit network. The computational results present savings in travelers’ journey time and operators’ costs, demonstrating the potential benefits of collaboration between multimodal transit systems and SAMS. (Pre-print)

Bio: Max Ng is a PhD Candidate in transportation engineering at Northwestern University. His research interests are public transit, smart mobility, network modeling, and data analytics. He is currently collaborating with Argonne National Laboratory on the project “Redesigning Multimodal Transit Systems with Shared Fleet Mobility Services”, with previous experience in rail decarbonization and freight resilience. Before joining Northwestern, Max obtained his Bachelor’s degree in Civil Engineering at the University of Hong Kong and worked in the MTR Corporation, a Hong Kong railway company. He is a chartered civil engineer in the United Kingdom and also the Institution of Civil Engineers Representative for the USA – Midwest.

Adrian Hernandez –Locating and sizing refueling facilities on networks to support freight rail decarbonization

Abstract: The decarbonization of the US transportation sector presents a considerable challenge for the freight railroads, whose operations are energy intensive and networks span tens of thousands of track-miles. We present a sequential framework to address the network deployment and sizing of the refueling infrastructure required by alternative fuel propulsion technologies. This paper presents two optimization problems to locate and size (i.e., specify the capacity of) alternative fuel refueling/charging facilities on existing rail networks. To illustrate the framework, we consider the deployment of battery-electric locomotives and charging facilities on an aggregate network of three Class I railroads. Our framework runs extremely efficiently and provides interpretable results at every step of the process. This framework also offers significant modeling flexibility as components can be altered to address different deployment objectives or operational considerations.

Bio: Adrian Hernandez is a current Ph.D. Candidate in Northwestern University’s Transportation Program in the Department of Civil and Environmental Engineering. At Northwestern, he is advised by Prof. Pablo Durango-Cohen, and has been conducting research on the decarbonization of freight rail in the U.S. Adrian received his B.S. in Civil Engineering from Cornell University, where he researched algorithms for vehicle routing problems under the supervision of Prof. Samitha Samaranayake and Juan Carlos Martinez Mori. Adrian has been the recipient of the LSAMP Fellowship and is a current GEM Fellow. His academic interests lie in optimization, operations research, infrastructure management, transportation, and public policy.

Francisco Contreras –Measuring the Impacts of Flooding on Traffic Using Crowdsourced Data

Abstract: Gathering traffic data used to be a challenging task. However, the massive adoption of location-based services has made it possible to obtain traffic data for almost any location in the United States. This data could allow us to measure traffic impacts during extreme events, such as flooding. Nevertheless, a few challenges remain. Platforms that provide traffic data deliver multiple traffic variables, and it is unclear which ones better reflect traffic disruption. Second, traffic data presents high variation daily, complicating determining when a flood event has ended. This research proposes a methodology to establish when the effects of a flood event on traffic start and finish and explores the best traffic variables to measure these effects on traffic. The results indicate that “Number of Trips,” “Vehicle Miles Traveled,” and “Vehicle Hours Traveled ” better capture the disruption a flood causes on traffic. Finally, the proposed methodology to establish the start and end of the flood events yields reasonable results.

Bio: Francisco Contreras is a second-year Ph.D. student at the University of Colorado Boulder, advised by Professor Cristina Torres-Machi. Francisco obtained a bachelor’s degree in civil engineering at the Technical University Federico Santa Maria, Chile. His research focuses on transportation asset management, and is currently working on how to measure the resilience of different communities’ transportation networks to flood events.

Ilia Papakonstantinou –Coastal protection strategies to minimize disruption to the Abu Dhabi transportation network from inundation due to sea level rise

Abstract: As sea levels rise, there has been an increase in research focused on the protection of shoreline infrastructure and transportation systems that may increasingly suffer from permanent capacity and accessibility reduction. This paper focuses on Abu Dhabi, UAE, a city vulnerable to inundation due to its insular geography, aiming to identify protection strategies that will minimize transportation network delays. The model considers hydrodynamic interactions and traffic assignment. The results show some shoreline portions are critical, and their protection leads to less congestion, while there are combinations of shoreline protection that worsen the congestion levels. The results also show that in some cases, the marginal effects of protecting one precinct may yield a better reduction of congestion than multiple other precincts. This research can provide a general framework for the protection of transportation infrastructure against sea level rise.

Bio: Ilia has received her Ph.D. in Transportation Planning & Engineering from New York University, and holds a degree in Electrical and Computer Engineering from the National Technical University of Athens, as well as an MSc in Engineering Systems and Management from Masdar Institute of Science and Technology, working on Operations Research. She currently works as a postdoctoral associate with the team of Prof. Samer Madanat at New York University Abu Dhabi. Her research focuses on highway infrastructure protection against sea level rise, as transportation network interactions in cases of inundation can lead to severe disruptions that cause considerable delays. Ilia is also interested in topics relevant to electric vehicles such as the electrification of bus systems.

Wang Chen –Eco-efficiency assessment of long-life asphalt pavement technologies

Abstract: Driven by climate change and other sustainable development pressures, long-life asphalt pavement (LLAP) technology has been recognized as a sustainable solution to reduce environmental burdens. This paper presented a comprehensive eco-efficiency assessment for six promising long-life asphalt pavement (LLAP) designs as a structural system, including the full-depth (FD), the thick asphalt layer (TAL), inverted (Inv), enhanced semi-rigid base (Esemi), composite (CP) and traditional semi-rigid base (Tsemi) asphalt pavement. Results showed that under the baseline design scenario, the Ive, ESemi and CP perform best in terms of greenhouse gas (GHG) emissions, total embodied energy (TEE), and life-cycle cost (LCC), respectively. In most cases of the scenario and uncertainty analyses, ESemi and Ive were the most eco-efficient solutions, although CP may dominate ESemi for a longer design life, whereas the other three designs were dominated by one or more of the abovementioned designs. Among all life-cycle activities, the raw material phase was proved to be the most critical source of GHG (54.2~79.4%), TEE (41.8~54.9%), and LCC (72.0~72.9%). Striking a balance between the content of cement-based and asphalt structural layers is the key to achieving sustainable pavement.

Bio: Wang Chen, currently a visiting scholar in the Department of Civil and Environmental Engineering at University of Waterloo, will be employed as a post-doctoral fellow at Risk-Informed Life-Cycle Infrastructure Engineering (RILCIE), Toronto Metropolitan University. With the cooperation of Dr. Arnold Yuan and Dr. Ningyuan Li, his research focuses on highway infrastructure system engineering management, particularly in life-cycle costing, life-cycle environmental assessment, sustainability-informed engineering design, data-driven deterioration modelling, maintenance & rehabilitation optimization and decision-making. Currently, he is working on incorporating deep uncertainties associated with emerging construction technologies and materials into pavement asset management for climate change mitigation; while assisting on a research project focused on developing a collaborative deterioration modelling platform for municipal infrastructure assets covering bridge, pavements, and sewer systems.

Deepak Benny – Transportation infrastructure finance: Challenges associated with state transportation agencies in identifying matching funds

Abstract: The Infrastructure Investment and Jobs Act has brought enormous opportunities for states in upgrading the existing infrastructure and to invest in future-ready transportation systems. On the other hand, the increased demand for matching fund (cost sharing from states) requirements has brought in elevated financial stress for State DOTs. The agencies thus need to identify alternative financing mechanisms to relieve the pressure from legislative budgetary constraints. Such approaches also help in bridge financing needs and to accelerate project implementation. Various smart innovative finance tools are engaged by State DOTs to meet the requirements. The strategies to be engaged and preparedness required while applying for federal grants are also discussed. The study identifies the challenges faced by agencies and various tools that can be engaged for availing matching funds. The trends across various State DOTs in matching strategies adopted are also discussed.

Bio: Deepak Benny is a PhD student in transportation engineering at Purdue University with a focus on Infrastructure finance. Deepak carries experience in planning, design and construction of infrastructure assets including roads, ports, marine terminals, industrial SEZs, the built environment and other transportation infrastructure. He has worked with international consulting firms based in Singapore and India, and has varied project experience across Southeast Asia, India, Middle East, and Europe. He holds a Masters’ Degree in Sustainable Engineering-Marine Technology from University of Glasgow, UK and Master of Business Administration (MBA) from Indian Institute of Management (IIM) Calicut, India.

Priyadarshan Patil – Railroad network electrification using symmetric traffic assignment

Abstract: We consider a budget-constrained rail network electrification problem with changes in energy usage, operations, and long-term maintenance costs. In particular, we consider that freight flows on such a network form a user equilibrium. Interactions between electric and diesel trains on the same corridor are represented with nonseparable link performance functions that nevertheless have a symmetric Jacobian. This bi-level problem is solved for the North American railroad network using a genetic algorithm, incorporating domain-specific insights to reduce the solution space. We analyze solution characteristics and decision-making implications. Results show that broad connectivity would be beneficial for the most impact. Increasing demand shifts electrified corridors toward the more populous east and gulf coasts, while increased operational costs results in the electrification of routes through mountainous terrains.

Bio: Priyadarshan Patil is an Applied Scientist at Amazon.com Inc., working with their network planning teams. He has a Ph.D. in Operations Research from The University of Texas at Austin, focusing on traffic assignment models and solution algorithms. Priyadarshan obtained his Bachelor’s degree in Civil Engineering at the Indian Institute of Technology Madras, and Master’s degree in Transportation Engineering from the University of Texas at Austin. His research focuses on network optimization, electrification and clean energy, and resilience.

Max Ng – Trading off Energy Storage and Payload – An Analytical Model for Freight Train Configuration

Abstract: We present a model to determine the optimal number of energy storage tender cars in a freight train with alternative fuel technologies (e.g., battery-electric locomotives) for decarbonization. For a given market and energy technology, the model captures the tradeoffs between inventory carrying costs associated with trip times, including delays, and ordering costs associated with train dispatch and operation (energy, amortized equipment and labor costs). We derive results for capital cost allocation based on either trip distance or time. The former leads to constant costs per order, whereas the latter to ordering costs that are functions of trip times. To illustrate the framework, we find the optimal number of battery-electric energy tender cars in freight markets from the 2019 Carload Waybill Sample for U.S. freight rail traffic collected by the Surface Transportation Board. The results display heterogeneity in optimal configurations with lighter, yet more time-sensitive shipments (e.g., intermodal) utilizing more battery tender cars. For heavier commodities (e.g., coal) with lower holding costs, single battery tender car configurations are optimal. The results also show that the optimal train configurations are sensitive to delays associated with recharging or swapping tender cars.

Bio: Max Ng is a Ph.D. Candidate in transportation engineering at Northwestern University. Max obtained his Bachelor’s degree in civil engineering at the University of Hong Kong. His research focuses on smart mobility, network analysis, and data analytics, with projects in rail decarbonization and freight resilience. Before joining Northwestern, he worked in MTR, a Hong Kong railway company, and is a chartered civil engineer in the UK and Hong Kong.

Meredith Raymer – An Area-wide Network Fundamental Diagram for Bicycles

Abstract: Recent work has incorporated the interactions between bicycles and cars into 3D network fundamental diagrams (NFDs). However, in these relationships the complete area on which the bike can travel has not been explicitly considered. While at some points in the network, bikes may have dedicated lanes or off-street paths, sharrows (shared car and bike lanes) allow a bike to traverse a wider area. By making use of the definitions of flow and density developed for pedestrians moving in areas, a bicycle NFD that accounts for the variation in width across a bike network is developed. As different car-bike interactions can occur between hard, soft and no separation environments, this work allows for these to be included in the derivation of NFDs through the area available for travel.

Bio: Meredith Raymer is a third-year PhD candidate at Northwestern University advised by Professor Hani Mahmassani. Her research interests are multi-modal urban networks and bridging gaps between urban planning and transportation engineering fields.

Recording

Sania Seilabi – Managing Dedicated Lanes for Connected and Autonomous Vehicles to Address Bottleneck Congestion Considering Morning Peak Commuter Departure Choices

Abstract: Connected and autonomous vehicles (CAVs) provide a valuable opportunity to address the growing traffic congestion in urban areas. In this context, the concept of the dedicated lane is proposed to increase mobility in the transition era, with a mixed fleet of human-driven vehicles (HDVs) and CAVs, toward a fully CAV fleet. This study investigates the impacts of CAV-dedicated lane(s) on traffic congestion during the morning peak period in a highway bottleneck. First, the user equilibrium condition is formulated as a linear program with complementarity constraints. Then, the solution existence and properties in terms of departure rates and queuing delays are demonstrated. It is proved that in any time interval, a CAV-dedicated lane has less queuing delay compared to a general-purpose lane. The system-optimal condition is formulated to obtain the minimum system cost, including total queuing delay, early and late arrival costs, by deriving the optimal number of lanes and tolling scheme. Our computational experiments suggest that the high CAV market penetration leads to less social inequity between CAV and HDV commuters. Further, it is shown that CAV technological advancement, which leads to further increased CAVL capacity, can significantly improve traffic mobility with an almost similar effect as a tolling scheme.

Bio: Sania Seilabi a National Science Foundation (NSF) eFellows postdoctoral fellow at the University at Buffalo (UB)’s Civil Engineering Department. Sania received her Ph.D. from Purdue University’s School of Civil Engineering, where she served as a research assistant at the USDOT Center for Connected & Automated Transportation (CCAT). She received her Bachelor of Science in Civil Engineering and Master’s degree in Civil Engineering with a specialization in Earthquake Engineering from Sharif University of Technology, Tehran, Iran. Sania has conducted research in a variety of areas related to advanced transportation mobility technologies, such as connected and autonomous vehicles (CAVs), and their implications for developing sustainable transportation infrastructure design and management.

Hamed Mehranfar – Optimal intervention planning of the railway infrastructure

Abstract: Determining the intervention requirements to be included in an intervention program requires the consideration of the assets, interventions, and the complex relationships between interventions, intervention costs and the service provided in a railway network. Digitalization has the potential to enable railway infrastructure managers to determine the time, location, and type of interventions more efficiently and effectively. This presentation will provide an overview of the possible methodologies to improve different aspects of the railway intervention planning process in a digital environment.

Bio: Hamed Mehranfar is a doctoral candidate in the chair of infrastructure management at the Institute of Construction and Infrastructure Management (IBI) at ETH Zurich. Hamed obtained his Master’s degree in railway transportation engineering with a focus on railway operations management from Iran University of Science and Technology (IUST) in 2019. Before joining ETH, he collected experience in applications of operations research in the railway industry by working in a railway transportation company from 2019-2020. Hamed’s research focuses on developing different tools for enabling asset managers to make better decisions.

Recording

Ajay Baniya – Durability Assessment of Externally Bonded Fiber-Reinforced Polymer (FRP) Composite Repairs in Bridges

Abstract: Although CFRP composites have been extensively used to rehabilitate many deficient bridges, data warranting their long-term performance is lacking. Current durability testing of CFRP composites involves accelerated conditioning as a part of material specification requirements to ensure that they maintain mechanical and physical properties during service life. However, without field data, relating accelerated conditioning test data to real-time outdoor exposure is not reliable. Work conducted at the University of Delaware in the early 2000s resulted in the application of externally bonded CFRP to strengthen a girder of a publicly owned steel bridge in the State of Delaware. As such, this bridge offers a unique opportunity to study CFRP durability over a time span of well over two decades. This report provides information on the durability of CFRP composites installed on bridge Br 1-704 in Newark, Delaware (USA). Field evaluation and laboratory testing of CFRP samples collected from several girders were employed to investigate CFRP degradation and bond quality. The results indicate that after more than two decades of service life, the condition of CFRP in Bridge 1-704 was found functional but the current Environmental reduction factor to compute the flexural strength and modulus per ACI40.2R (2017) must be revisited to ensure the structural integrity of a strengthened bridge.

Bio: Ajay Baniya is a registered Professional Engineer (PE) and Structural Engineer (SE) in the State of Nevada. He has more than 5 years of experience as a Civil/Structural Engineer, and about 3 years of research experience as a Graduate Research Assistant. He completed his MS in Structural Engineering from the University of Texas at Arlington in 2017, where he studied the performance of Light-Frame wooden structures (LFWSs) subjected to combined wind and flood hazards. Currently, he is working on his Ph.D. candidacy exam at the University of Delaware, and he has been doing research on the durability performance of CFRP composites subjected to environmental conditioning and fatigue stresses for more than 2 decades. The CFRP composites were used to strengthen two public steel bridges located in the State of Delaware. His research interests include CFRP composites, machine learning, fatigue stresses in steel bridges, seismic engineering, and building information modeling. Aside from research and engineering, he loves being outdoors, playing soccer, working out, and camping.

Sandra Milev – Polysulfide Elastomers as Self-healing Sealants for Transportation Infrastructure

Abstract: As transportation agencies largely neglect joint sealant maintenance in concrete pavements, water infiltration into the deteriorated joint and subgrade can result in base softening, erosion, and faulting, posing a significant financial burden on taxpayers. To improve the durability performance of concrete pavement sealants, an elastomer with self-healing ability was designed by incorporating dynamic disulfide bonds. The sealant was prepared using commercially available liquid polysulfides and epoxy resin. Sealant performance and self-healing were optimized by tuning the molecular weight of polysulfide oligomers, the stoichiometric ratio of epoxy to thiol, and catalyst content. Preliminary mechanical tests under ambient conditions show that reducing the stoichiometric ratio from 1 to 0.9 and increasing the molecular weight of polysulfides from 4000 to 4500 g/mol increased the ultimate elongation of self-healed samples from 391% to 612%. The addition of 1% of catalyst further increased the elongation of self-healed samples to 828%, achieving 87% healing efficiency (measured as a percentage of recovered elongation). Analysis of data on thermal properties is underway.

Bio: Sandra Milev is a research assistant and a Ph.D. candidate in the Department of Civil Engineering at University of Delaware. Her research is related to the durability of FRP composites and self-healing polymers for application in construction.

Recording

Amir Fard – Infrastructure network maintenance planning using a deep reinforcement learning approach

Abstract: Civil infrastructure networks are subject to gradual aging as they are exposed to various structural and operational deterioration mechanisms throughout their life cycle. Maintaining these systems in good condition is crucial for any country’s economic growth and social development. Identifying the optimal intervention strategies for infrastructure asset management is a complex sequential decision-making problem, and due to the sheer number of decision alternatives, it is a computationally challenging task that can be intractable for large-scale networks, especially, when the intervention decisions associated with the network components are dependent due to functional or economic dependencies such as budget constraints. In this research, I propose an infrastructure network maintenance planning framework based on Markov decision processes (MDP) and using deep reinforcement learning (DRL) approach to manage large discrete action space. As a case study, the flushing program of a sewer network for a municipality in Ontario is optimized for a defined planning horizon. To evaluate the performance of the framework, the results will be compared to other optimization algorithms such as genetic algorithm.

Bio: Amir Fard is a PhD candidate in the Civil Engineering Department at Toronto Metropolitan University, Canada. He holds bachelor’s and master’s degrees in Aerospace Engineering from Sharif University of Technology.  Under the supervision of Dr. Arnold Yuan, his research focuses on Risk-informed infrastructure life-cycle planning. Currently, he is working on developing a multi-agent deep reinforcement learning framework for adaptive inspection and maintenance plan optimization for infrastructure networks.

Sophia Rupp – Improving the Durability of Externally Bonded Carbon Fiber Reinforced Polymer (CFRP) Composites with Fiber Anchors

Abstract: Externally bonded carbon fiber-reinforced polymer (CFRP) composites have become a material of choice for repair and strengthening of deficient concrete bridges. While CFRP offers significant advantages over traditional repair materials (e.g., steel) – such as higher strength-to-weight ratio, improved durability, and accelerated construction – the bond between CFRP and concrete was found to deteriorate when exposed to harsh environments, particularly moisture. To address this problem and promote enhanced long-term performance of CFRP in concrete bridges, this research investigates the effect of fiber anchors on the strength retention of concrete beams reinforced with externally bonded CFRP following accelerated conditioning. The presentation covers preliminary work performed to assess the behavior of concrete beams with anchored CFRP. Moreover, test methodology, the selected accelerated conditioning protocol, and scope of the ongoing durability study will be addressed.

Bio: Sophia Rupp is currently a civil engineering Master’s student with a concentration in structural engineering at the University of Delaware. Her Master’s research is focused on studying advanced applied materials for existing concrete bridge rehabilitation and strengthening. She also obtained her Bachelor’s in civil engineering at the University of Delaware while assisting on a research project focused on comparing different US State’s load rating procedures for bridges or culverts with missing or incomplete as-built information. She hopes to either continue working with the current bridge rehabilitation and strengthening material she is studying or continue research of new materials in the industry after she graduates from the University of Delaware.

Recording

Hannah Power – Passive strain sensing for structural health monitoring using retroreflective sheeting materials

Abstract: Retroreflective sheeting materials (RRSM) are used for various applications in engineering, but primarily for traffic signs and pavement markings. There are several ASTM standard types of RRSM that have required values of retroreflection to ensure safe usage. Retroreflectivity (RR) is the portion of light returned to the light source measured in candelas per lux per square meter, as measured using a retroreflectometer. It is theorized that as load is applied to RRSM, the retroreflection will change and have a reasonably linear relationship to the material’s strain (ε), thus, opening the possibility for using RRSM material as a passive strain sensor for structural health monitoring that is low cost, practical, and innovative. A total of ten materials, manufactured by Avery Dennison and 3M, were loaded in tension and their strain and retroreflectivity were measured at 4.45 and 8.90 Newton (one and two pound) load intervals. Specimens were subjected to three loading and unloading cycles, and their sensitivity (retroreflectivity divided by strain) was calculated to determine their feasibility for use as passive strain sensors. Results show that the retroreflectivity of RRSM do change when subjected to load, and that certain materials demonstrate a reasonably linear relation to strain. Furthermore, others do not return to their virgin retroreflectivity but degrade through the course of repeated loading. The ten materials were also subjected to strength tests to determine their failure stresses and strain to failure. Four of the materials have the highest retroreflective sensitivity to strain and have high failure stresses and strains and are the most likely to perform well as passive strain sensors.

Bio: Hannah Power is a PhD candidate in the Civil Engineering Department at the University of Delaware. She works with Dr. Tripp Shenton researching retroreflective sheeting materials and their potential use as passive strain sensors for structural health monitoring. Hannah holds a bachelor’s and master’s degree in Civil Engineering from Villanova University, conferred in 2016 and 2020, respectively. Before coming to the University of Delaware, Hannah worked for Michael Baker International for four years in the greater Philadelphia region designing and rehabilitating bridges.

Yuto Nakazato – Optimal Daily Repair Schedule for Road Network

Abstract: Many studies have been presented to optimize long-term repair policies that specify facilities to repair in the road network within a year. However, we cannot implement the repair policy without a repair schedule that identifies the facilities to repair on each day. This study targets on the optimal daily repair schedule for the road network when facilities need repair in the year is already decided by the long-term repair policy. In terms of the network cost, we consider the repair costs of the road network that has economies of the scale, and the user costs that changes according to the capacity of the network and the traffic demand. The problem is formulated as a mixed integer linear programming problem (MIP), and the network user costs and the network repair cost are expressed in linear form so that optimal solution can be easily computed using a general optimizer. Numerical results show that the optimal daily repair schedule can be calculated in acceptable time and that the optimal daily repair schedule can reduce both the user cost and the repair cost of the road network compared to the randomly decided daily repair schedule.

Bio: Yuto Nakazato was originally born in Tokyo but spent his lifetime till high-school in Shanghai, China, which allowed him to be able speak both Chinese and Japanese. Yuto returned to Japan for his bachelor’s degree and has been studying at Tohoku University in Japan for 7 years. Now, as a 1st year Ph.D. student, Yuto studies infrastructure management under Professor Mizutani’s supervision. His main research interests lie in optimization of repair policies and repair schedules for large-scale infrastructure systems.

Recording

David Zani – Do cost-benefit analyses lead to good projects? Our first insights.

Abstract: Existing research and data show that both cost and benefit estimates for infrastructure projects are inaccurate in many parts of the world. Costs are systematically underestimated, and benefits are systematically overestimated. Cost-benefit-analyses are therefore inefficient as a decision-making aid, potentially leading to suboptimal decision making. The following research will 1) show if and to what extent this problem is present in public-sector infrastructure projects, 2) show the potential advantages of gathering and managing cost and benefit data, and 3) suggest an organization of project data to provide such advantages. This talk specifically will discuss some of our first insights into this topic from project data gathered thus far (part of work package 1).

Bio: David Zani is a PhD Candidate at the Infrastructure Management Group at ETH Zurich, the Swiss Federal Institute of Technology. He is working together with Prof. Bryan Adey to further research on the use of cost-benefit analyses. In 2021, David obtained his master’s degree from ETH Zurich in Spatial Development and Infrastructure Systems. Before studying in Zürich, he attended the University of Texas at Austin, where he received his bachelor’s degree in Civil Engineering from the Cockrell School of Engineering in 2016. He has held various internships, including engineering and risk consulting, and construction contracting and management.

Jing Yu – Considering Environmental Repercussions in Pavement Design: An Input-Output Model with Substitution

Abstract: In industrial manufacturing, a final product is usually built up from various raw materials and sub-assemblies. If some of these materials are substituted for each other, then the product designer is forced to make a choice. In this paper, we develop a computational framework for finding the best ecological friendly combination of alternatives in product design that minimizes the total greenhouse gas emission. To this end, we first develop a product structure framework to keep track of all the required materials for assembling the final product and represent it as a matrix form. We leverage the input-output (I-O) life cycle assessment (LCA) method to analyze the production framework, which results in a redundant system of linear equations whose solution set tracks all possible combinations of alternatives that fulfill the final demand. The product design problem is then formulated as a linear program (LP), whose objective is to minimize the production cost as well as the total emissions. We show that the proposed framework has three advantages: (1) it scales well to the total number of alternatives; (2) it can be efficiently solved via LP algorithms; (3) it allows conducting sensitivity analysis analytically. Eventually, we test our method on a numerical example of designing a pavement with three layers (surface, subbase, and base), each of which can be built using 2 or 3 alternative materials. We first solve the LP and then perform sensitivity analysis on multiple system parameters (e.g., production costs and emission rates). The analytic results of sensitivity analysis are validated by directly comparing with the numerical solutions after changing the parameters.

Bio: Jing Yu is currently pursuing her Ph.D. in Civil and Environmental Engineering at Northwestern University. She holds a Bachelor of Science degree in Aviation Engineering and a Master of Science degree in Civil Engineering. Under the supervision of Professor Pablo Durango-Cohen, her research focuses on the life cycle assessment and environmental design of transportation systems. She develops a framework to support the environmentally friendly product design, which can analyze the life cycle greenhouse gas emissions of pavement construction and provide insights of material selection under different scenarios.