Nils Boysen, Dirk Briskorn, Rea Röntgen, Michael Dienstknecht
Among the most crucial organizational challenges of free-floating carsharing is the question how to cope with regional demand imbalance. Because users are allowed to leave a rented car anywhere in the service district, it regularly occurs that too many cars are left behind in low-demand regions whereas other regions face a demand surplus. In this paper, we consider a countermeasure that has been overlooked by previous research: an optimization-based matching of carsharing supply and demand that not only addresses the profit promised by the current matches but also targets future demand imbalance. To account for such imbalances, we define regional demand levels that specify the projected number of requested cars per region and aim to reduce the deviations of the regions’ actual car supply from these target levels. We present exact polynomial-time algorithms for this extended matching task that are suitable for real-time application on large carsharing platforms. In an extensive computational study, we compare optimization-based matching approaches with and without the consideration of demand imbalance and benchmark them with the status quo, the individual choice of carsharing users among available cars. Based on generated data with considerable demand variation among regions, our results indicate a clear advantage of our novel matching approach. In a further study based on a large carsharing data set, however, the proof of concept fails because the real-world regions are cut according to geographical characteristics instead of demand variation. To successfully relieve the strains of demand imbalance, our novel matching task thus requires a properly partitioned service district and reliable forecasts of the carsharing demands.Funding: This work was supported by the Deutsche Forschungsgemeinschaft [Grants BO 3148/8-1 and BR 3873/10-1].Supplemental Material: The online appendix is available at https://doi.org/10.1287/trsc.2022.0067 .
{"title":"Matching vs. Individual Choice: How to Counter Regional Imbalance of Carsharing Demand","authors":"Nils Boysen, Dirk Briskorn, Rea Röntgen, Michael Dienstknecht","doi":"10.1287/trsc.2022.0067","DOIUrl":"https://doi.org/10.1287/trsc.2022.0067","url":null,"abstract":"Among the most crucial organizational challenges of free-floating carsharing is the question how to cope with regional demand imbalance. Because users are allowed to leave a rented car anywhere in the service district, it regularly occurs that too many cars are left behind in low-demand regions whereas other regions face a demand surplus. In this paper, we consider a countermeasure that has been overlooked by previous research: an optimization-based matching of carsharing supply and demand that not only addresses the profit promised by the current matches but also targets future demand imbalance. To account for such imbalances, we define regional demand levels that specify the projected number of requested cars per region and aim to reduce the deviations of the regions’ actual car supply from these target levels. We present exact polynomial-time algorithms for this extended matching task that are suitable for real-time application on large carsharing platforms. In an extensive computational study, we compare optimization-based matching approaches with and without the consideration of demand imbalance and benchmark them with the status quo, the individual choice of carsharing users among available cars. Based on generated data with considerable demand variation among regions, our results indicate a clear advantage of our novel matching approach. In a further study based on a large carsharing data set, however, the proof of concept fails because the real-world regions are cut according to geographical characteristics instead of demand variation. To successfully relieve the strains of demand imbalance, our novel matching task thus requires a properly partitioned service district and reliable forecasts of the carsharing demands.Funding: This work was supported by the Deutsche Forschungsgemeinschaft [Grants BO 3148/8-1 and BR 3873/10-1].Supplemental Material: The online appendix is available at https://doi.org/10.1287/trsc.2022.0067 .","PeriodicalId":51202,"journal":{"name":"Transportation Science","volume":"80 S8","pages":""},"PeriodicalIF":4.6,"publicationDate":"2023-11-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"138495574","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Christian Müller, Jochen Gönsch, Matthias Soppert, Claudius Steinhardt
Free-floating vehicle sharing systems such as car or bike sharing systems offer customers the flexibility to pick up and drop off vehicles at any location within the business area and, thus, have become a popular type of urban mobility. However, this flexibility has the drawback that vehicles tend to accumulate at locations with low demand. To counter these imbalances, pricing has proven to be an effective and cost-efficient means. The fact that customers use mobile applications, combined with the fact that providers know the exact location of each vehicle in real-time, provides new opportunities for dynamic pricing. In this context of modern vehicle sharing systems, we develop a profit-maximizing dynamic pricing approach that is built on adopting the concept of customer-centricity. Customer-centric dynamic pricing here means that, whenever a customer opens the provider’s mobile application to rent a vehicle, the price optimization incorporates the customer’s location as well as disaggregated choice behavior to precisely capture the effect of price and walking distance to the available vehicles on the customer’s probability for choosing a vehicle. Two other features characterize the approach. It is origin-based, that is, prices are differentiated by location and time of rental start, which reflects the real-world situation where the rental destination is usually unknown. Further, the approach is anticipative, using a stochastic dynamic program to foresee the effect of current decisions on future vehicle locations, rentals, and profits. We propose an approximate dynamic programming-based solution approach with nonparametric value function approximation. It allows direct application in practice, because historical data can readily be used and main parameters can be precomputed such that the online pricing problem becomes tractable. Extensive numerical studies, including a case study based on Share Now data, demonstrate that our approach increases profits by up to 8% compared with existing approaches from the literature. History: This paper has been accepted for the Transportation Science Special Issue on 2021 TSL Workshop: Supply and Demand Interplay in Transport and Logistics. Supplemental Material: The e-companion is available at https://doi.org/10.1287/trsc.2021.0524 .
{"title":"Customer-Centric Dynamic Pricing for Free-Floating Vehicle Sharing Systems","authors":"Christian Müller, Jochen Gönsch, Matthias Soppert, Claudius Steinhardt","doi":"10.1287/trsc.2021.0524","DOIUrl":"https://doi.org/10.1287/trsc.2021.0524","url":null,"abstract":"Free-floating vehicle sharing systems such as car or bike sharing systems offer customers the flexibility to pick up and drop off vehicles at any location within the business area and, thus, have become a popular type of urban mobility. However, this flexibility has the drawback that vehicles tend to accumulate at locations with low demand. To counter these imbalances, pricing has proven to be an effective and cost-efficient means. The fact that customers use mobile applications, combined with the fact that providers know the exact location of each vehicle in real-time, provides new opportunities for dynamic pricing. In this context of modern vehicle sharing systems, we develop a profit-maximizing dynamic pricing approach that is built on adopting the concept of customer-centricity. Customer-centric dynamic pricing here means that, whenever a customer opens the provider’s mobile application to rent a vehicle, the price optimization incorporates the customer’s location as well as disaggregated choice behavior to precisely capture the effect of price and walking distance to the available vehicles on the customer’s probability for choosing a vehicle. Two other features characterize the approach. It is origin-based, that is, prices are differentiated by location and time of rental start, which reflects the real-world situation where the rental destination is usually unknown. Further, the approach is anticipative, using a stochastic dynamic program to foresee the effect of current decisions on future vehicle locations, rentals, and profits. We propose an approximate dynamic programming-based solution approach with nonparametric value function approximation. It allows direct application in practice, because historical data can readily be used and main parameters can be precomputed such that the online pricing problem becomes tractable. Extensive numerical studies, including a case study based on Share Now data, demonstrate that our approach increases profits by up to 8% compared with existing approaches from the literature. History: This paper has been accepted for the Transportation Science Special Issue on 2021 TSL Workshop: Supply and Demand Interplay in Transport and Logistics. Supplemental Material: The e-companion is available at https://doi.org/10.1287/trsc.2021.0524 .","PeriodicalId":51202,"journal":{"name":"Transportation Science","volume":"20 2","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-11-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135634848","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Louis Bouvier, Guillaume Dalle, Axel Parmentier, Thibaut Vidal
This paper is the fruit of a partnership with Renault. Their reverse logistic requires solving a continent-scale multiattribute inventory routing problem (IRP). With an average of 30 commodities, 16 depots, and 600 customers spread across a continent, our instances are orders of magnitude larger than those in the literature. Existing algorithms do not scale, so we propose a large neighborhood search (LNS). To make it work, (1) we generalize existing split delivery vehicle routing problems and IRP neighborhoods to this context, (2) we turn a state-of-the-art matheuristic for medium-scale IRP into a large neighborhood, and (3) we introduce two novel perturbations: the reinsertion of a customer and that of a commodity into the IRP solution. We also derive a new lower bound based on a flow relaxation. In order to stimulate the research on large-scale IRP, we introduce a library of industrial instances. We benchmark our algorithms on these instances and make our code open source. Extensive numerical experiments highlight the relevance of each component of our LNS. Funding: This work was supported by Renault Group. Supplemental Material: The online appendix is available at https://doi.org/10.1287/trsc.2022.0342 .
{"title":"Solving a Continent-Scale Inventory Routing Problem at Renault","authors":"Louis Bouvier, Guillaume Dalle, Axel Parmentier, Thibaut Vidal","doi":"10.1287/trsc.2022.0342","DOIUrl":"https://doi.org/10.1287/trsc.2022.0342","url":null,"abstract":"This paper is the fruit of a partnership with Renault. Their reverse logistic requires solving a continent-scale multiattribute inventory routing problem (IRP). With an average of 30 commodities, 16 depots, and 600 customers spread across a continent, our instances are orders of magnitude larger than those in the literature. Existing algorithms do not scale, so we propose a large neighborhood search (LNS). To make it work, (1) we generalize existing split delivery vehicle routing problems and IRP neighborhoods to this context, (2) we turn a state-of-the-art matheuristic for medium-scale IRP into a large neighborhood, and (3) we introduce two novel perturbations: the reinsertion of a customer and that of a commodity into the IRP solution. We also derive a new lower bound based on a flow relaxation. In order to stimulate the research on large-scale IRP, we introduce a library of industrial instances. We benchmark our algorithms on these instances and make our code open source. Extensive numerical experiments highlight the relevance of each component of our LNS. Funding: This work was supported by Renault Group. Supplemental Material: The online appendix is available at https://doi.org/10.1287/trsc.2022.0342 .","PeriodicalId":51202,"journal":{"name":"Transportation Science","volume":"72 3","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-10-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135765682","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Operators who deploy large fleets of electric vehicles often face a challenging charge scheduling problem. Specifically, time-ineffective recharging operations limit the profitability of charging during service operations such that operators recharge vehicles off duty at a central depot. Here, high investment cost and grid capacity limit available charging infrastructure such that operators need to schedule charging operations to keep the fleet operational. In this context, flexible service operations, that is, allowing delayed or expedited vehicle departures, can potentially increase charger utilization. Beyond this, jointly scheduling charging and service operations promises operational cost savings through better utilization of time-of-use energy tariffs and carefully crafted charging schedules designed to minimize battery wear. Against this background, we study the resulting joint charging and service operations scheduling problem accounting for battery degradation, nonlinear charging, and time-of-use energy tariffs. We propose an exact branch-and-price algorithm, leveraging a custom branching rule and a primal heuristic to remain efficient during the branch-and-bound phase. Moreover, we develop an exact labeling algorithm for our pricing problem, constituting a resource-constrained shortest path problem that considers variable energy prices and nonlinear charging operations. We benchmark our algorithm in a comprehensive numerical study and show that it can solve problem instances of realistic size with computational times below one hour, thus enabling its application in practice. Additionally, we analyze the benefit of jointly scheduling charging and service operations. We find that our integrated approach lowers the amount of charging infrastructure required by up to 57% besides enabling operational cost savings of up to 5%. Funding: This work was supported by the German Federal Ministry for Economic Affairs and Energy [Grant 01MV21020B]. Supplemental Material: The online appendix is available at https://doi.org/10.1287/trsc.2022.0272 .
{"title":"Electric Vehicle Charge Scheduling with Flexible Service Operations","authors":"Patrick S. Klein, Maximilian Schiffer","doi":"10.1287/trsc.2022.0272","DOIUrl":"https://doi.org/10.1287/trsc.2022.0272","url":null,"abstract":"Operators who deploy large fleets of electric vehicles often face a challenging charge scheduling problem. Specifically, time-ineffective recharging operations limit the profitability of charging during service operations such that operators recharge vehicles off duty at a central depot. Here, high investment cost and grid capacity limit available charging infrastructure such that operators need to schedule charging operations to keep the fleet operational. In this context, flexible service operations, that is, allowing delayed or expedited vehicle departures, can potentially increase charger utilization. Beyond this, jointly scheduling charging and service operations promises operational cost savings through better utilization of time-of-use energy tariffs and carefully crafted charging schedules designed to minimize battery wear. Against this background, we study the resulting joint charging and service operations scheduling problem accounting for battery degradation, nonlinear charging, and time-of-use energy tariffs. We propose an exact branch-and-price algorithm, leveraging a custom branching rule and a primal heuristic to remain efficient during the branch-and-bound phase. Moreover, we develop an exact labeling algorithm for our pricing problem, constituting a resource-constrained shortest path problem that considers variable energy prices and nonlinear charging operations. We benchmark our algorithm in a comprehensive numerical study and show that it can solve problem instances of realistic size with computational times below one hour, thus enabling its application in practice. Additionally, we analyze the benefit of jointly scheduling charging and service operations. We find that our integrated approach lowers the amount of charging infrastructure required by up to 57% besides enabling operational cost savings of up to 5%. Funding: This work was supported by the German Federal Ministry for Economic Affairs and Energy [Grant 01MV21020B]. Supplemental Material: The online appendix is available at https://doi.org/10.1287/trsc.2022.0272 .","PeriodicalId":51202,"journal":{"name":"Transportation Science","volume":"42 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-10-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"136381588","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Sami Serkan Özarık, Paulo da Costa, Alexandre M. Florio
In the context of the Amazon Last-Mile Routing Research Challenge, this paper presents a machine-learning framework for optimizing last-mile delivery routes. Contrary to most routing problems where an objective function is clearly defined, in the real-world setting considered in the challenge, an objective is not explicitly specified and must be inferred from data. Leveraging techniques from machine learning and classical traveling salesman problem heuristics, we propose a “pool and select” algorithm to prescribe high-quality last-mile delivery sequences. In the pooling phase, we exploit structural knowledge acquired from data, such as common entry and exit regions observed in training routes. In the selection phase, we predict the scores of candidate sequences with a high-dimensional, pretrained, and regularized regression model. The score prediction model, which includes a large number of predictor variables such as sequence duration, compliance with time windows, earliness, lateness, and structural similarity to training data, displays good prediction accuracy and guides the selection of efficient delivery sequences. Overall, the framework is able to prescribe competitive delivery routes, as measured on out-of-sample routes across several data sets. Given that desired characteristics of high-quality sequences are learned and not assumed, the proposed framework is expected to generalize well to last-mile applications beyond those immediately foreseen in the challenge. Moreover, the method requires less than three seconds to prescribe a sequence given an instance and, thus, is suitable for very large-scale applications. History: This paper has been accepted for the Transportation Science Special Issue on Machine Learning Methods and Applications in Large-Scale Route Planning Problems. Funding: This research was funded by The Dutch Research Council (NWO) Data2Move project under [Grant 628.009.013] and the European Union’s Horizon 2020 research and innovation programme under the Marie Sklodowska-Curie [Grant 754462]. Supplemental Material: The online appendix is available at https://doi.org/10.1287/trsc.2022.0029 .
{"title":"Machine Learning for Data-Driven Last-Mile Delivery Optimization","authors":"Sami Serkan Özarık, Paulo da Costa, Alexandre M. Florio","doi":"10.1287/trsc.2022.0029","DOIUrl":"https://doi.org/10.1287/trsc.2022.0029","url":null,"abstract":"In the context of the Amazon Last-Mile Routing Research Challenge, this paper presents a machine-learning framework for optimizing last-mile delivery routes. Contrary to most routing problems where an objective function is clearly defined, in the real-world setting considered in the challenge, an objective is not explicitly specified and must be inferred from data. Leveraging techniques from machine learning and classical traveling salesman problem heuristics, we propose a “pool and select” algorithm to prescribe high-quality last-mile delivery sequences. In the pooling phase, we exploit structural knowledge acquired from data, such as common entry and exit regions observed in training routes. In the selection phase, we predict the scores of candidate sequences with a high-dimensional, pretrained, and regularized regression model. The score prediction model, which includes a large number of predictor variables such as sequence duration, compliance with time windows, earliness, lateness, and structural similarity to training data, displays good prediction accuracy and guides the selection of efficient delivery sequences. Overall, the framework is able to prescribe competitive delivery routes, as measured on out-of-sample routes across several data sets. Given that desired characteristics of high-quality sequences are learned and not assumed, the proposed framework is expected to generalize well to last-mile applications beyond those immediately foreseen in the challenge. Moreover, the method requires less than three seconds to prescribe a sequence given an instance and, thus, is suitable for very large-scale applications. History: This paper has been accepted for the Transportation Science Special Issue on Machine Learning Methods and Applications in Large-Scale Route Planning Problems. Funding: This research was funded by The Dutch Research Council (NWO) Data2Move project under [Grant 628.009.013] and the European Union’s Horizon 2020 research and innovation programme under the Marie Sklodowska-Curie [Grant 754462]. Supplemental Material: The online appendix is available at https://doi.org/10.1287/trsc.2022.0029 .","PeriodicalId":51202,"journal":{"name":"Transportation Science","volume":"8 4","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-10-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135273087","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Using a parsimonious model, this paper analyzes a dockless bike-sharing (DLB) service that competes with walking and a generic motorized mode. The DLB operator chooses a fleet size and a fare schedule that dictate the level of service (LOS) as measured by the access time or the walking time taken to reach the nearest bike location. The market equilibrium is formulated as a solution to a nonlinear equation system over which three counterfactual design problems are defined to maximize (i) profit, (ii) ridership, or (iii) social welfare. The model is calibrated with empirical data collected in Chengdu, China, and all three counterfactual designs are tested against the status quo. We show the LOS of a DLB system is subject to rapidly diminishing returns to the investment on the fleet. Thus, under the monopoly setting considered herein, the current fleet cap set by Chengdu can be cut by up to three quarters even when the DLB operator aims to maximize ridership. This indicates the city’s fleet cap decision might have been misguided by the prevailing conditions of a competitive yet highly inefficient market. For a regulator seeking to influence the DLB operator for social good, the choice of policy instruments depends on the operator’s objective. When the operator focuses on profit, limiting fare is much more effective than limiting fleet size. If, instead, it aims to grow market share, then setting a limit on fleet size becomes a dominant strategy. We also show, both analytically and numerically, that the ability to achieve a stable LOS with a low rebalancing frequency is critical to profitability. A lower rebalancing frequency always rewards users with cheaper fares and better LOS even for a profit-maximizing operator. Funding: This research was partially supported by the U.S. National Science Foundation [Grant CMMI 1922665] and the National Natural Science Foundation of China [Grant 71971044]. Supplemental Material: The online appendix is available at https://doi.org/10.1287/trsc.2022.0304 .
{"title":"How Many Are Too Many? Analyzing Dockless Bike-Sharing Systems with a Parsimonious Model","authors":"Hongyu Zheng, Kenan Zhang, Yu (Marco) Nie, Pengyu Yan, Yuan Qu","doi":"10.1287/trsc.2022.0304","DOIUrl":"https://doi.org/10.1287/trsc.2022.0304","url":null,"abstract":"Using a parsimonious model, this paper analyzes a dockless bike-sharing (DLB) service that competes with walking and a generic motorized mode. The DLB operator chooses a fleet size and a fare schedule that dictate the level of service (LOS) as measured by the access time or the walking time taken to reach the nearest bike location. The market equilibrium is formulated as a solution to a nonlinear equation system over which three counterfactual design problems are defined to maximize (i) profit, (ii) ridership, or (iii) social welfare. The model is calibrated with empirical data collected in Chengdu, China, and all three counterfactual designs are tested against the status quo. We show the LOS of a DLB system is subject to rapidly diminishing returns to the investment on the fleet. Thus, under the monopoly setting considered herein, the current fleet cap set by Chengdu can be cut by up to three quarters even when the DLB operator aims to maximize ridership. This indicates the city’s fleet cap decision might have been misguided by the prevailing conditions of a competitive yet highly inefficient market. For a regulator seeking to influence the DLB operator for social good, the choice of policy instruments depends on the operator’s objective. When the operator focuses on profit, limiting fare is much more effective than limiting fleet size. If, instead, it aims to grow market share, then setting a limit on fleet size becomes a dominant strategy. We also show, both analytically and numerically, that the ability to achieve a stable LOS with a low rebalancing frequency is critical to profitability. A lower rebalancing frequency always rewards users with cheaper fares and better LOS even for a profit-maximizing operator. Funding: This research was partially supported by the U.S. National Science Foundation [Grant CMMI 1922665] and the National Natural Science Foundation of China [Grant 71971044]. Supplemental Material: The online appendix is available at https://doi.org/10.1287/trsc.2022.0304 .","PeriodicalId":51202,"journal":{"name":"Transportation Science","volume":"30 10","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-10-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135366349","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The distribution of passenger vehicles is a complex task and a high cost factor for automotive original equipment manufacturers (OEMs). On the way from the production plant to the customer, vehicles travel long distances on different carriers, such as ships, trains, and trucks. To save costs, OEMs and logistics service providers aim to maximize their loading capacities. Modern auto carriers are extremely flexible. Individual platforms can be rotated, extended, or combined to accommodate vehicles of different shapes and weights and to nest them in a way that makes the best use of the available space. In practice, finding feasible combinations is done with the help of simple heuristics or based on personal experience. In research, most papers that deal with auto carrier loading focus on route or cost optimization. Only a rough approximation of the loading subproblem is considered. In this paper, we present two different methodologies to approximate realistic load factors considering the flexibility of modern auto carriers and their height, length, and weight constraints. Based on our industry partner’s process, the vehicle distribution follows a first in, first out principle. For the first approach, we formulate the problem as a mixed integer, quadratically constrained assignment problem. The second approach considers the problem as a two-dimensional nesting problem with irregular shapes. We perform computational experiments using real-world data from a large German automaker to validate and compare both models with each other and with an approximate model adapted from the literature. The simulation results for the first approach show that, on average, for 9.37% of all auto carriers, it is possible to load an additional vehicle compared with the current industry solution. This translates to 1.36% less total costs. The performance of the nesting approach is slightly worse, but as it turns out, it is well-suited to check load combinations for feasibility.
{"title":"Load Factor Optimization for the Auto Carrier Loading Problem","authors":"Christian Jäck, Jochen Gönsch, Hans Dörmann-Osuna","doi":"10.1287/trsc.2022.0373","DOIUrl":"https://doi.org/10.1287/trsc.2022.0373","url":null,"abstract":"The distribution of passenger vehicles is a complex task and a high cost factor for automotive original equipment manufacturers (OEMs). On the way from the production plant to the customer, vehicles travel long distances on different carriers, such as ships, trains, and trucks. To save costs, OEMs and logistics service providers aim to maximize their loading capacities. Modern auto carriers are extremely flexible. Individual platforms can be rotated, extended, or combined to accommodate vehicles of different shapes and weights and to nest them in a way that makes the best use of the available space. In practice, finding feasible combinations is done with the help of simple heuristics or based on personal experience. In research, most papers that deal with auto carrier loading focus on route or cost optimization. Only a rough approximation of the loading subproblem is considered. In this paper, we present two different methodologies to approximate realistic load factors considering the flexibility of modern auto carriers and their height, length, and weight constraints. Based on our industry partner’s process, the vehicle distribution follows a first in, first out principle. For the first approach, we formulate the problem as a mixed integer, quadratically constrained assignment problem. The second approach considers the problem as a two-dimensional nesting problem with irregular shapes. We perform computational experiments using real-world data from a large German automaker to validate and compare both models with each other and with an approximate model adapted from the literature. The simulation results for the first approach show that, on average, for 9.37% of all auto carriers, it is possible to load an additional vehicle compared with the current industry solution. This translates to 1.36% less total costs. The performance of the nesting approach is slightly worse, but as it turns out, it is well-suited to check load combinations for feasibility.","PeriodicalId":51202,"journal":{"name":"Transportation Science","volume":"31 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-10-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"136033168","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Xiaoxu Chen, Zhanhong Cheng, Jian Gang Jin, Martin Trépanier, Lijun Sun
Accurate forecasting of bus travel time and its uncertainty is critical to service quality and operation of transit systems: it can help passengers make informed decisions on departure time, route choice, and even transport mode choice, and it also support transit operators on tasks such as crew/vehicle scheduling and timetabling. However, most existing approaches in bus travel time forecasting are based on deterministic models that provide only point estimation. To this end, we develop in this paper a Bayesian probabilistic model for forecasting bus travel time and estimated time of arrival (ETA). To characterize the strong dependencies/interactions between consecutive buses, we concatenate the link travel time vectors and the headway vector from a pair of two adjacent buses as a new augmented variable and model it with a mixture of constrained multivariate Gaussian distributions. This approach can naturally capture the interactions between adjacent buses (e.g., correlated speed and smooth variation of headway), handle missing values in data, and depict the multimodality in bus travel time distributions. Next, we assume different periods in a day share the same set of Gaussian components, and we use time-varying mixing coefficients to characterize the systematic temporal variations in bus operation. For model inference, we develop an efficient Markov chain Monte Carlo (MCMC) algorithm to obtain the posterior distributions of model parameters and make probabilistic forecasting. We test the proposed model using the data from two bus lines in Guangzhou, China. Results show that our approach significantly outperforms baseline models that overlook bus-to-bus interactions, in terms of both predictive means and distributions. Besides forecasting, the parameters of the proposed model contain rich information for understanding/improving the bus service, for example, analyzing link travel time and headway correlation using covariance matrices and understanding time-varying patterns of bus fleet operation from the mixing coefficients. Funding: This research is supported in part by the Fonds de Recherche du Quebec-Societe et Culture (FRQSC) under the NSFC-FRQSC Research Program on Smart Cities and Big Data, the Canadian Statistical Sciences Institute (CANSSI) Collaborative Research Teams grants, and the Natural Sciences and Engineering Research Council (NSERC) of Canada. X. Chen acknowledges funding support from the China Scholarship Council (CSC). Supplemental Material: The e-companion is available at https://doi.org/10.1287/trsc.2022.0214 .
{"title":"Probabilistic Forecasting of Bus Travel Time with a Bayesian Gaussian Mixture Model","authors":"Xiaoxu Chen, Zhanhong Cheng, Jian Gang Jin, Martin Trépanier, Lijun Sun","doi":"10.1287/trsc.2022.0214","DOIUrl":"https://doi.org/10.1287/trsc.2022.0214","url":null,"abstract":"Accurate forecasting of bus travel time and its uncertainty is critical to service quality and operation of transit systems: it can help passengers make informed decisions on departure time, route choice, and even transport mode choice, and it also support transit operators on tasks such as crew/vehicle scheduling and timetabling. However, most existing approaches in bus travel time forecasting are based on deterministic models that provide only point estimation. To this end, we develop in this paper a Bayesian probabilistic model for forecasting bus travel time and estimated time of arrival (ETA). To characterize the strong dependencies/interactions between consecutive buses, we concatenate the link travel time vectors and the headway vector from a pair of two adjacent buses as a new augmented variable and model it with a mixture of constrained multivariate Gaussian distributions. This approach can naturally capture the interactions between adjacent buses (e.g., correlated speed and smooth variation of headway), handle missing values in data, and depict the multimodality in bus travel time distributions. Next, we assume different periods in a day share the same set of Gaussian components, and we use time-varying mixing coefficients to characterize the systematic temporal variations in bus operation. For model inference, we develop an efficient Markov chain Monte Carlo (MCMC) algorithm to obtain the posterior distributions of model parameters and make probabilistic forecasting. We test the proposed model using the data from two bus lines in Guangzhou, China. Results show that our approach significantly outperforms baseline models that overlook bus-to-bus interactions, in terms of both predictive means and distributions. Besides forecasting, the parameters of the proposed model contain rich information for understanding/improving the bus service, for example, analyzing link travel time and headway correlation using covariance matrices and understanding time-varying patterns of bus fleet operation from the mixing coefficients. Funding: This research is supported in part by the Fonds de Recherche du Quebec-Societe et Culture (FRQSC) under the NSFC-FRQSC Research Program on Smart Cities and Big Data, the Canadian Statistical Sciences Institute (CANSSI) Collaborative Research Teams grants, and the Natural Sciences and Engineering Research Council (NSERC) of Canada. X. Chen acknowledges funding support from the China Scholarship Council (CSC). Supplemental Material: The e-companion is available at https://doi.org/10.1287/trsc.2022.0214 .","PeriodicalId":51202,"journal":{"name":"Transportation Science","volume":"120 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-10-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"136295697","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Sihan Wang, Roberto Baldacci, Yang Yu, Yu Zhang, Jiafu Tang, Xinggang Luo, Wei Sun
Motivated by the worldwide development of shared mobility, we investigate a vehicle routing problem with time windows and deadlines called the first-mile ridesharing problem (FMRSP). The FMRSP involves routing a fleet of vehicles, each servicing customers within specific time windows. The FMRSP generalizes the well-known vehicle routing problem with time windows (VRPTW), additionally imposing that each vehicle route arrives at the destination before the earliest deadline associated with the set of customers served by the route. The FMRSP is also related to the VRPTW and release dates, where in addition to time window constraints, a release date is associated with each customer defining the earliest time that the order is available to leave the depot for delivery. For the FMRSP, we present an exact method based on a branch-price-and-cut (BPC) algorithm combining state-of-the-art techniques and an innovative pricing algorithm. The pricing algorithm is based on a bidirectional bucket graph-based labeling algorithm, in which the backward extension of a label is computed in a constant time. Effective dominance rules used to speed up the computation are also described. Extensive computational studies demonstrate that our proposed BPC algorithm can solve optimality-modified Solomon benchmark instances involving up to 100 customers and real-world instances involving up to 290 customers. Funding: This research was supported by the National Natural Science Foundation of China [Grants 71831003, 72171043, 71831006, and 71901180]. Supplemental Material: The online appendix is available at https://doi.org/10.1287/trsc.2022.0139 .
{"title":"An Exact Method for a First-Mile Ridesharing Problem","authors":"Sihan Wang, Roberto Baldacci, Yang Yu, Yu Zhang, Jiafu Tang, Xinggang Luo, Wei Sun","doi":"10.1287/trsc.2022.0139","DOIUrl":"https://doi.org/10.1287/trsc.2022.0139","url":null,"abstract":"Motivated by the worldwide development of shared mobility, we investigate a vehicle routing problem with time windows and deadlines called the first-mile ridesharing problem (FMRSP). The FMRSP involves routing a fleet of vehicles, each servicing customers within specific time windows. The FMRSP generalizes the well-known vehicle routing problem with time windows (VRPTW), additionally imposing that each vehicle route arrives at the destination before the earliest deadline associated with the set of customers served by the route. The FMRSP is also related to the VRPTW and release dates, where in addition to time window constraints, a release date is associated with each customer defining the earliest time that the order is available to leave the depot for delivery. For the FMRSP, we present an exact method based on a branch-price-and-cut (BPC) algorithm combining state-of-the-art techniques and an innovative pricing algorithm. The pricing algorithm is based on a bidirectional bucket graph-based labeling algorithm, in which the backward extension of a label is computed in a constant time. Effective dominance rules used to speed up the computation are also described. Extensive computational studies demonstrate that our proposed BPC algorithm can solve optimality-modified Solomon benchmark instances involving up to 100 customers and real-world instances involving up to 290 customers. Funding: This research was supported by the National Natural Science Foundation of China [Grants 71831003, 72171043, 71831006, and 71901180]. Supplemental Material: The online appendix is available at https://doi.org/10.1287/trsc.2022.0139 .","PeriodicalId":51202,"journal":{"name":"Transportation Science","volume":"41 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-09-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135244255","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
This paper focuses on the estimation and compression of ride demand from origin-destination (OD) trip event data. By representing the OD event data as a three-way tensor (origin, destination, and time), we model the data as a Poisson process with an intensity tensor that can be decomposed according to a Tucker decomposition. We establish and justify a specific form of nonnegative Tucker-like tensor decomposition that represents OD demand via K latent origin spatial factors and K latent destination spatial factors. We then provide a computational and memory efficient algorithm for performing this decomposition and demonstrate its use for real-time compression and estimation of OD ride demand. Two case studies based on New York City (NYC) taxi and Washington DC (DC) taxi were implemented. Results from the case studies demonstrate the applicability of the proposed method in data compression and short-term forecast for ride demand. Furthermore, we found that the learned latent spatial factors are interpretable and localized to specific areas for both NYC and DC cases. Hence, this method can be used to understand OD trip data through latent spatial factors and be used to identify spatio-temporal patterns for OD trip and travel demand generation mechanism in general. Funding: This work was supported by the U.S. Department of Transportation [UTC/NCST] and the U.S. National Science Foundation [Grant DMS 1712996]. Supplemental Material: The online appendix is available at https://doi.org/10.1287/trsc.2022.0101 .
{"title":"Understanding Origin-Destination Ride Demand with Interpretable and Scalable Nonnegative Tensor Decomposition","authors":"Xiaoyue Li, Ran Sun, James Sharpnack, Yueyue Fan","doi":"10.1287/trsc.2022.0101","DOIUrl":"https://doi.org/10.1287/trsc.2022.0101","url":null,"abstract":"This paper focuses on the estimation and compression of ride demand from origin-destination (OD) trip event data. By representing the OD event data as a three-way tensor (origin, destination, and time), we model the data as a Poisson process with an intensity tensor that can be decomposed according to a Tucker decomposition. We establish and justify a specific form of nonnegative Tucker-like tensor decomposition that represents OD demand via K latent origin spatial factors and K latent destination spatial factors. We then provide a computational and memory efficient algorithm for performing this decomposition and demonstrate its use for real-time compression and estimation of OD ride demand. Two case studies based on New York City (NYC) taxi and Washington DC (DC) taxi were implemented. Results from the case studies demonstrate the applicability of the proposed method in data compression and short-term forecast for ride demand. Furthermore, we found that the learned latent spatial factors are interpretable and localized to specific areas for both NYC and DC cases. Hence, this method can be used to understand OD trip data through latent spatial factors and be used to identify spatio-temporal patterns for OD trip and travel demand generation mechanism in general. Funding: This work was supported by the U.S. Department of Transportation [UTC/NCST] and the U.S. National Science Foundation [Grant DMS 1712996]. Supplemental Material: The online appendix is available at https://doi.org/10.1287/trsc.2022.0101 .","PeriodicalId":51202,"journal":{"name":"Transportation Science","volume":"111 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-09-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"136312976","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：