Search

In this paper we describe a container loading software package for a pulp and paper manufacturer. The program takes, as input, orders for paper products in sheets or reels. As output it produces precise order quantities, and stowage plans for shipping containers. The program includes a heuristic algorithm for two-dimensional stowing of reels of a common diameter, a bin packing algorithm for combining reels into stacks and a two-dimensional algorithm for stowing pallets of various sizes. Since the order sizes are usually greater than one container load all algorithms are designed for the multi-container case. The program is currently in use and has been effective in several decision situations including (a) good sizes for reel diameters (b) efficient stowage patterns, and (c) production quantities that are compatible with cost efficient shipping. It also reduces the time taken to produce stowage and shipping documents.

To manage the risk of insurance companies, a reinsurance transaction is among the myriad risk management mechanisms the top ranked choice. In this paper, we study the design of optimal reinsurance contracts within a risk measure minimization framework and subject to the Vajda condition. The Vajda condition requires the reinsurer to take an increasing proportion of the loss when it increases and therefore imposes constraints on the indemnity function. The distortion-risk-measure-based objective function is very generic, and allows for, for example, an objective to minimize the risk-adjusted value of the insurer’s liability, and for heterogeneous beliefs regarding the loss distribution by the insurer and reinsurer. Under a mild condition, we propose a backward-forward optimization method that is based on a marginal indemnity function formulation. To show the applicability and simplicity of our strategy, we provide three concrete examples with the Value-at-Risk: one with the risk-adjusted value of the insurer’s liability, one with an objective function that follows from imposing Pareto optimality, and one with heterogeneous beliefs. We conclude this paper with an empirical application with Danish fire insurance losses and the Value-at-Risk and the Tail Value-at-Risk.

The maximum weighted submatrix coverage problem is a recently introduced problem with applications in data mining. It is concerned with selecting K submatrices of a given numerical matrix such that the sum of the matrix-entries, which occur in at least one of the selected submatrices, is maximized. In the paper introducing the problem, a problem-specific constraint programming approach was developed and embedded in a large neighborhood-search to obtain a heuristic. A compact integer linear programming formulation was also presented, but deemed inefficient due to its size. In this paper, we introduce new integer linear programming formulations for the problem, one of them is based on Benders decomposition. The obtained Benders decomposition-based formulation has a nice combinatorial structure, i.e., there is no need to solve linear programs to separate Benders cuts. We present preprocessing procedures and valid inequalities for all formulations. We also develop a greedy randomized adaptive search procedure for the problem, which is enhanced with a local search. A computational study using the instances from literature is done to evaluate the effectiveness of our new approaches. Our algorithms manage to find improved primal solutions for ten out of 17 real-world instances, and optimality is proven for two real-world instances. Moreover, for over 700 of 1617 large-scale synthetic instances, our algorithms find improved primal solutions compared to the heuristics from the literature.

This paper proposes a novel matheuristic algorithm for solving the Production Routing Problem (PRP). The PRP is a hard-to-solve combinatorial optimization problem with numerous practical applications in the field of freight transportation, logistics and supply chain management. A manufacturer is responsible for determining production decisions, as well as the timing and the quantity of replenishment services offered to a set of geographically dispersed customers over a multi-period time horizon. The problem calls for jointly optimizing the production, inventory, distribution and routing decisions. The paper provides a novel two-commodity flow formulation and proposes a two-phase infeasible space matheuristic algorithm for solving the examined problem. The first phase deals with a relaxation of the problem to construct production-distribution plans. In the second phase, these are completed with routing information and optimized via a local search framework which oscillates between the feasible and the infeasible solution space. The framework is equipped with mixed integer programming components for restoring infeasibility and for diversifying the conducted search. Computational experiments demonstrate that the infeasibility space exploration significantly contributes to the quality of the final solutions. The results obtained by the proposed matheuristic out-match the results of state-of-the-art approaches. More specifically, 594 and 55 new best solutions out of 1440 and 90 instances of two well-established benchmark data sets of small-medium and large instances are reported, respectively.

This paper examines the high-dimensional dependence modelling problem in the context of project risk assessment. As the dimension of uncertain performance units (i.e., itemized costs and activity times) in a project increases, specifying a feasible correlation matrix and eliciting relevant pair-wise information, either from historical data or with expert judgement, becomes practically unattainable or simply not economical. This paper presents a factor-driven dependence elicitation and modelling framework with scalability to large-scale project risks. The multi-factor association model (MFAM) accounts for hierarchical relationships of multiple association factors and provides a closed-form solution to a complete and mathematically consistent correlation matrix. Augmented with the structured association (SA) technique for systematic identification of hierarchical association factors, the MFAM offers additional flexibility of utilizing the minimum information available in standardized, ubiquitous project plans (e.g., work breakdown structure, resource allocation, or risk register), while preserving the computational efficiency and the scalability to high dimensional project risks. Numerical applications and simulation experiments show that the MFAM, further combined with extended analytics (i.e., parameter calibration and optimization), provides credible risk assessments (with accuracy comparable to full-scale simulation) and further enhances the realism of dealing with high-dimensional project risks utilizing all relevant information.

Metaheuristics are widely employed to solve hard optimization problems, like vehicle routing problems (VRP), for which exact solution methods are impractical. In particular, local search-based metaheuristics have been successfully applied to the capacitated VRP (CVRP). The CVRP aims at defining the minimum-cost delivery routes for a given set of identical vehicles since each vehicle only travels one route and there is a single (central) depot. The best metaheuristics to the CVRP avoid getting stuck in local optima by embedding specific hill-climbing mechanisms such as diversification strategies into the solution methods. This paper introduces a hybridization of a novel adaptive version of Iterated Local Search with Path-Relinking (AILS-PR) to the CVRP. The major contribution of this paper is an automatic mechanism to control the diversity step of the metaheuristic to allow it to escape from local optima. The results of experiments with 100 benchmark CVPR instances show that AILS-PR outperformed the state-of-the-art CVRP metaheuristics.

Wildfires can pose a major threat to the secure operation of power networks. Chile, California, and Australia have suffered from recent wildfires that have induced considerable power supply cuts. Further, as power systems move to a significant integration of variable renewable energy sources, successfully managing the impact of wildfires on the power supply can become even more challenging due to the joint uncertainty in wildfire trajectories and the power injections from wind and solar farms. Motivated by this, this paper develops a practical decision-support approach that concatenates a stochastic wildfire simulation method with an attacker-defender model that aims to find a worst-case realization for (i) transmission line and generator contingencies, out of those that can potentially be affected by a given wildfire scenario, and for (ii) wind and solar power trajectories, based on a max-min structure where the inner min problem represents a best adaptive response on generator dispatch actions. Further, this paper proposes an evaluation framework to assess the power supply security of various power system topology configurations, under the assumption of limited transmission switching capabilities, and based on the simulation of several wildfire evolution scenarios. Extensive computational experiments are carried out on two representations of the Chilean power network with up to 278 buses, showing the practical effectiveness of the proposed approach for enhancing wildfire resilience in highly renewable power systems.

In this paper, a strategic decision making model for the sustainable development of marine renewable energy is proposed, and a specific application to the United Kingdom (UK) is demonstrated. As an island nation the UK benefits from significant marine energy potential which is providing an increasing contribution to UK's renewable energy portfolio. The paper investigates the question of how decision makers can be aided to reach a decision on which types of marine renewable energy projects should be chosen for development given that strategic energy planning is subject to a number of uncertain parameters and multiple sustainability objectives. In this context, the contribution of this paper lies in the combination of renewable energy portfolio selection and the application of multi-objective methods. Interval coefficient goal programming models are adopted in order to address the impreciseness and uncertainty associated with the goals and coefficients of the models in the context of renewable energy selection. The potential renewable energy projects are clustered in order to aid the decision making process and preferential weight sensitivity methods are employed. Conclusions are drawn for optimistic and pessimistic scenarios in the context of UK marine renewable energy planning.

This paper surveys the recent attempts, both from the machine learning and operations research communities, at leveraging machine learning to solve combinatorial optimization problems. Given the hard nature of these problems, state-of-the-art algorithms rely on handcrafted heuristics for making decisions that are otherwise too expensive to compute or mathematically not well defined. Thus, machine learning looks like a natural candidate to make such decisions in a more principled and optimized way. We advocate for pushing further the integration of machine learning and combinatorial optimization and detail a methodology to do so. A main point of the paper is seeing generic optimization problems as data points and inquiring what is the relevant distribution of problems to use for learning on a given task.

Attended home delivery describes the delivery of goods by e-grocers or e-tailers to customers within an agreed time window. Because customers expect narrow time windows, offering such services may lead to expensive fulfillment operations. This has led to research on how to influence customers’ bookings using time window pricing or slotting. In this paper, we reconsider the problem of demand management through dynamic pricing for attended home delivery services. This problem is usually modeled as a stochastic dynamic program, but even small instances cannot be solved to optimality due to the curses of dimensionality. The major challenges consist of finding feasible time windows for an incoming customer, estimating the opportunity costs, i.e., the future monetary loss due to accepting a booking, and optimizing the time window prices in real time. In this paper, we propose a route-based approximate dynamic programming approach to tackle these challenges. The approach carefully combines and partially extends state-of-the-art methods in attended home delivery, dynamic pricing, and dynamic vehicle routing. In an extensive simulation study, we compare its performance with state-of-the-art benchmark heuristics. The results indicate a superior performance of our approach in terms of both profit and number of customers served.

A preventive maintenance scheduling problem is studied on behalf of generation companies (GENCOs) with natural gas power plants, while taking into account their signed natural gas contracts and the opportunities of purchasing and selling natural gas in the spot market. This paper considers the uncertain prices of both natural gas and electricity in the spot market, and proposes a multistage stochastic mixed integer programming (MSMIP) model seeking the optimal operations regarding maintenance outage scheduling and natural gas trading. Large-scale MSMIP problems suffer not only the curse of dimensionality, but also computational difficulties with both discrete and continuous variables at each stage. To this respect, this paper leverages the progressive hedging algorithm based on scenario-based decomposition to solve large MSMIP problems. The solutions obtained from the algorithm exhibit promising quality under our numerical studies. Due to the independence among all the subproblems after the decomposition, the algorithm is amenable to parallel computing, which leads to faster convergence as demonstrated in the numerical results. Computational experiments also show that it is beneficial to use MSMIP while considering both maintenance planning and natural gas contracting. In addition, the results also indicate the GENCOs with a larger number of small generators perform better than those with a smaller number of big generators.

Hyper-heuristics have emerged as a way to raise the level of generality of search techniques for computational search problems. This is in contrast to many approaches, which represent customised methods for a single problem domain or a narrow class of problem instances. The term hyper-heuristic was defined in the early 2000s as a heuristic to choose heuristics, but the idea of designing high-level heuristic methodologies can be traced back to the early 1960s. The current state-of-the-art in hyper-heuristic research comprises a set of methods that are broadly concerned with intelligently selecting or generating a suitable heuristic for a given situation. Hyper-heuristics can be considered as search methods that operate on lower-level heuristics or heuristic components, and can be categorised into two main classes: heuristic selection and heuristic generation. Here we will focus on the first of these two categories, selection hyper-heuristics. This paper gives a brief history of this emerging area, reviews contemporary selection hyper-heuristic literature, and discusses recent selection hyper-heuristic frameworks. In addition, the existing classification of selection hyper-heuristics is extended, in order to reflect the nature of the challenges faced in contemporary research. Unlike the survey on hyper-heuristics published in 2013, this paper focuses only on selection hyper-heuristics and presents critical discussion, current research trends and directions for future research.

To reduce congestion, environmental damage, and negative health impact in large urban areas plenty of novel concepts for last-mile distribution have been innovated in the recent years. The concept treated in this paper are mobile parcel lockers that are able to change their locations during the day, either autonomously or moved by a human driver. By relocating lockers their reach towards addressees also varying their whereabouts over the day can be increased. This paper optimizes the changing locations of lockers, such that customers are at some time during the planning horizon within a predefined range of their designated locker. Our aim is to minimize the locker fleet when satisfying all customers. We formulate the resulting mobile locker location problem and provide suited exact solution procedures. To asses the potential whether mobile lockers are a promising last-mile concept, worth the investment required to develop it to a market-ready solution, we benchmark the necessary fleet size of mobile lockers with the required number of their stationary counterparts. Our results show that considerable reductions are possible.

This paper studies how appointment cancellations affect scheduling strategies in outpatient healthcare clinics. While cancellation rates in outpatient clinics have been reported to be as high as 27%, cancelled appointments are often ignored, or grouped with no-shows in healthcare scheduling models. We find that there exists a value of total demand that, when calculated over a scheduling horizon, marks the boundary between where cancellations hurt or help a clinic. We refer to that value as the switch point. Up to the switch point, clinics can achieve a greater reward when patients do not cancel. However, for values of expected total demand greater than the switch point, the clinic reward is reduced if more patients retain (do not cancel) appointments. To assist us in evaluating the switch point, we construct a mixed-integer nonlinear programming model to solve a multi-day outpatient scheduling problem. The model accounts for both inter-day (appointment day) and intra-day (appointment time slot) scheduling decisions, while balancing service benefits against service costs. We include probabilities of no-show and cancellation, which allows us to discuss how cancellations affect scheduling decisions through the switch point. The knowledge of the switch point allows a clinic to understand when appointment no-shows and cancellations negatively affect clinic service, and can assist the clinic in determining the number of patients for which it is committed to provide service, i.e., its panel size. In this paper, we discuss methodologies for calculating the switch point, and discuss its sensitivity to model parameters.

The problem studied in this paper is the multi-depot open vehicle routing problem, which has the following two differences in relation to the classical vehicle routing problem: there are several depots; the vehicles do not return to the depot after delivering the goods to the customers, i.e., the end of the route is not the starting point. There are many practical applications for this problem, however the great majority of the studies have only addressed the open vehicle routing problem with a single depot. In this paper, we present an iterated local search algorithm, in which the moves performed during the local search are recalled and this historical search information is then used to define the moves executed inside the perturbation procedures. Therefore, it is recorded the number of times that each customer is moved during the local search. Since this information is continuously updated and changes in each iteration, the search is driven to potentially better regions of the solution space, and increases the chance of avoiding cycling, even when using deterministic perturbations. The performance of this algorithm was tested using a large set of benchmark problems and was compared with other algorithms, and the results show that it is very competitive.

Shelter opening and evacuation of vulnerable populations are operations crucial to disaster response, which is one of the four phases of Disaster Operations Management (DOM). Optimisation has tried to capture some of the different issues related to shelter location and evacuation routing: several models have been developed over the years. However, they are still far from being fully comprehensive. The aim of this paper is to identify the current challenges in devising realistic and applicable optimisation models in the shelter location and evacuation routing context, with the ultimate goal of outlining a roadmap for future research in this topical area. A critical analysis of the most recent combined models is provided, including insights from the authors of the existing papers. The analysis highlights numerous gaps and research opportunities, such as the need for future optimisation models to involve stakeholders, include evacuee as well as system behaviour, be application-oriented rather than theoretical or model-driven, and interdisciplinary.

Many institutions have their own vehicle fleet and hired drivers that can be assigned to working trips to ensure that they all take place within the desired time periods. In these situations, it is important that vehicles and drivers are assigned to planned trips in an optimal and automated way. Trips are planned to occur over a given planning horizon and have a number of characteristics that have to be taken into account: a trip is characterized by a starting time, start and end locations, travelling distance and time, number of drivers needed for each assigned vehicle, number of passengers, and so on. This paper considers the problem of assigning both vehicles and drivers to a set of planned trips, having as objective the minimization of total cost. To properly address this problem it is necessary to consider a number of features, like the possibility of vehicles being shared between different trips if the timeframes and destinations are compatible. The mathematical model developed and presented in this paper takes all these features into account. The problem was motivated by a case study in the context of a public Brazilian university. This case study is also described and computational experiments using a general solver are reported. The general solver is capable of calculating optimal solutions in reasonable computational times, but problems where the fleet is mainly composed of small vehicles are more challenging. A matheuristic has been developed and tested, that seems to be competitive for these type of problems.

In surface mining operations, fleet management systems seek to make optimal decisions to handle material in two steps: path production optimization and real-time truck dispatching. This paper develops a multiple objective transportation model for real-time truck dispatching. The model addresses two major drawbacks of former models. The proposed model dispatches the trucks to destinations trying to simultaneously minimize shovel idle times, truck wait times, and deviations from the path production requirements established by the production optimization stage. To evaluate the performance of the proposed model, we developed a benchmark model based on the backbone of the most widely used fleet management system in the mining industry (Modular Mining DISPATCH). Afterward, we built a discrete event simulation model of the truck and shovel operation using an iron ore mine case study, implemented both of the dispatching models, and compared the results. The implementation of the models suggests that the multiple objective model developed in this paper is able to meet the production requirements of the operation using a fleet at 85% of the size of the deterministically calculated desired fleet. In addition, the model is able to meet the full capacity of the processing plants with a fleet of 30% less trucks than the desired fleet.

In conventional project management, the project manager is given a deadline and is responsible for making decisions assuming this deadline will not change. But in reality, project deadlines frequently change. Conventional project management addresses unanticipated changes in the project deadline with external change control processes. If this external process changes the deadline, the manager will replace the original deadline with the modified deadline. The manager is still suppose to assume this modified deadline is fixed in making project decisions. As this paper will show, recognizing deadline uncertainty can substantially improve the value of the manager’s decisions. Deadline uncertainty could be formally addressed in project management by making dramatic changes in conventional project management. But given the widespread use of project management, making such dramatic changes might not be feasible. This paper presents a much more easily implemented approach to incorporating deadline uncertainty into managerial decision making which makes no changes to conventional project management procedures.

Social media may play a critical role in disaster response by identifying needs in a shorter amount of time and thereby improving situational awareness. However, needs identified through social media initially have not been verified and some may be inaccurate. This can create a barrier to its use during disaster response decision making. Consequently, a key tradeoff between the timeliness and accuracy of social data for disaster response logistics planning must be assessed. This study aims to stimulate interest in a research agenda aimed at evaluating this tradeoff. A general framework for investigating whether it is worthwhile to act on user-generated data prior to its absolute verification in the context of disaster response logistics planning is presented. While the framework is applicable to a variety of logistics planning problems, this paper demonstrates its use via an application in mobile delivery of disaster response commodities. A case study motivated by the 2010 Haiti earthquake is developed using real social data, and is presented for use in conjunction with the framework. The types of insights that are possible via the framework are revealed through the small computational study included in the paper.

Community Operational Research (Community OR), and its disciplinary relation, Community-Based Operations Research, has an increasingly high profile within multiple domains that benefit from empirical and analytical approaches to problem solving. These domains are primarily concentrated within nonprofit services and local development. However, there are many other disciplines and application areas for which novel applications and extensions of Community OR could generate valuable insights. This paper identifies a number of these, distinguishing between ‘emerging trends’ (mostly in well-studied areas of operational research, management science and analytics) and ‘new frontiers’, which can be found in traditions not commonly oriented towards empirical and analytical methods for problem solving, where community-engaged decision modeling represents new ways of generating knowledge, policies and prescriptions. This paper will show how the exploration of emerging trends and new frontiers in Community OR can provide a basis for the development of innovative research agendas that can broaden the scope and impact of the decision sciences.

Electric Energy Storage Systems (EESS) have received an increased attention in recent years due to their important role in an active management of energy supply systems. Fueled by the increasing shares of intermittent Renewable Energy Sources (RES) in today's energy supply, balancing energy demand and energy supply over time becomes more and more challenging. EESS are recognized as a key technology to overcome this challenge by storing energy and converting it back when needed. Even though some EESS solutions are already available on the market, EESS suffer from technical limitations and entail high investment costs. Energy management is responsible for managing the operations of EESS and the interactions with the surrounding systems. An optimal energy management is an important precondition to ensure economic viability of EESS. This paper presents a systematic review of the literature on energy management for stationary EESS applications. The aim of the paper is to give a comprehensive overview of the literature in this field and to develop a conceptual framework that facilitates the structuring of research on the management of EESS and the identification of future research opportunities. The paper first introduces the methodology used for the literature review and then descriptively analyzes the identified publications. Afterwards, it outlines the proposed conceptual framework that provides a structure for categorizing the identified literature. Subsequently, we discuss the identified papers in light of the proposed framework. The paper then concludes with a discussion of future research opportunities.

The literature suggests that increasing stakeholder engagement has a positive impact on projects using discrete-event simulation in healthcare. This suggests projects should strive to involve the stakeholders in as much of the project as possible, through facilitated workshops. A notable gap in stakeholder involvement is the model coding stage, in which a conceptual model is turned into a discrete-event simulation model running on a computer. This paper investigates how and under what circumstances model coding might also be conducted in facilitated workshops, in particular through the use of the Business Process Model and Notation (BPMN) modelling standard. This work arose from a series of modelling projects with two hospitals, one in Italy and the other in the UK. The paper describes how BPMN can contribute, with a case in which model coding was achieved in a facilitated workshop and a second in which it was not but which highlights further barriers to this in some contexts. These barriers arise from the detail necessary for requisite modelling regarding (i) the level of complexity of the model and (ii) challenges in data access and analysis to populate the model. The relationship between the technical capabilities of tools available and the impact of these barriers is also discussed. We believe this is the first time that discrete-event simulation model coding in a facilitated workshop in healthcare has been described, and we provide a clear view of the further barriers. To indicate when facilitated model coding is currently achievable, we suggest a contextual matrix.

Local search (LS) is an essential module of most hybrid meta-heuristic evolutionary algorithms which are a major approach aimed to solve efficiently multi-objective optimization (MOO) problems. Furthermore, LS is specifically useful in many real-world applications where there is a need only to improve a current state of a system locally with limited computational budget and/or relying on computationally expensive process simulators. In these contexts, this paper proposes a new neighborhood-based iterative LS method, relying on first derivatives approximation and linear programming (LP), aiming to steer the search along any desired direction in the objectives space. The paper also leverages the directed local search (DS) method to constrained MOO problems. These methods are applied to the bi-objective (cost versus life cycle assessment-based environmental impact) optimization of drinking water production plants. The results obtained show that the proposed method constitutes a promising local search method which clearly outperforms the directed search approach.

This paper investigates the performance of the Yahoo crowd and experts in predicting the outcomes of matches in the World Cup in 2014. The analysis finds that the Yahoo crowd was statistically significantly better at predicting outcomes of matches than experts and very similar in performance to established betting odds. In addition, this paper finds that there was a statistically significant difference between the Yahoo crowd and a different crowd's performances, for the same task, suggesting that characteristics of the “crowd matter.” Finally, this paper finds that different crowdsourcing approaches apparently provide different results. Accordingly, it is important to specify the particular crowdsourcing approach, rather than simply “crowdsource.”

In this paper, the effect of behavioural analytics on short-term default predictions at the invoice level is addressed by answering a question that slightly diverges from the traditional probability of default definition: ‘What is the probability that this invoice will be paid within the next 30 days?’ Resultantly improving short-term liquidity planning accuracy and supporting financial management in companies. To provide a valid answer to the research question, a set of issues needs to be resolved, including identifying an appropriate data set, increasing the data predictive power, and creating and testing predictive models. Since the appropriate data set is not yet presented, we primarily focus on the first two issues: identifying appropriate data and raising its predictive power. In this paper, we propose to build predictive models upon a new data source from multiple companies, acquired by business partners' data sharing concept. Furthermore, we upgrade these data with behavioural analysis to test the assumption that the probability of default depends not only on payment capability but also on payment preparedness. The predictive power of shared invoice data and the effects of behavioural analysis are tested in a two-phase experiment: first, basic shared data are used to predict short-term invoice defaults, and in the second phase, the behavioural analysis results are included in the dataset. Lastly, the predictive models’ test results are compared. Both results are positive: the already high accuracy of models, build upon basic data is significantly upgraded in models, using the behaviour analysis extended data set.

Seriation aims at finding a linear order for a set of objects to reveal structural information which can be used for deriving data-driven decisions. It presents a difficult combinatorial optimization problem with its roots and applications in many fields including operations research. This paper focuses on a popular seriation problem which tries to find an order for a single set of objects that optimizes a given seriation criterion defined on one-mode two-way data, i.e., an object-by-object dissimilarity matrix. Over the years, members of different research communities have introduced many criteria and seriation methods for this problem. It is often not clear how different seriation criteria and methods relate to each other and which criterion or seriation method to use for a given application. These methods are representing tools for analytics and therefore are of theoretical and practical interest to the operations research community. The purpose of this paper is to provide a consistent overview of the most popular criteria and seriation methods and to present a comprehensive experimental study to compare their performance using artificial and a representative set of real-world datasets.

This paper considers dynamic booking control for a single-station car rental revenue management problem. Different from conventional airline revenue management, car rental revenue management needs to take into account not only the existing bookings but also the lengths of the existing rentals and the capacity flexibility via fleet shuttling, which yields a high-dimensional system state space. In this paper, we formulate the dynamic booking control problem as a discrete-time stochastic dynamic program over an infinite horizon. Such a model is computationally intractable. We propose a decomposition approach and develop two heuristics. The first heuristic is an approximate dynamic program (ADP) which approximates the value function using the value functions of the decomposed problems. The second heuristic is constructed directly from the optimal booking limits computed from the decomposed problems, which is more scalable compared to the ADP heuristic. Our numerical study suggests that the performances of both heuristics are close to optimum and significantly outperform the commonly used probabilistic non-linear programming (PNLP) heuristic in most of the instances. The dominant performance of our second heuristic is evidenced in a case study using sample data from a major car rental company in the UK.

As one of the most challenging combinatorial optimization problems in scheduling, the resource-constrained project scheduling problem (RCPSP) has attracted numerous scholars’ interest resulting in considerable research in the past few decades. However, most of these papers focused on the single objective RCPSP; only a few papers concentrated on the multi-objective resource-constrained project scheduling problems (MORCPSP). Inspired by a procedure called electromagnetism (EM), which can help a generic population-based evolutionary search algorithm to obtain good results for single objective RCPSP, in this paper we attempt to extend EM and integrate it into three reputable state-of-the-art multi-objective evolutionary algorithms (MOEAs) i.e. non-dominated sorting based multi-objective evolutionary algorithm (NSGA-II), strength Pareto evolutionary algorithm (SPEA2) and multi-objective evolutionary algorithm based on decomposition (MOEA/D), for MORCPSP. We aim to optimize makespan and total tardiness. Empirical analysis based on standard benchmark datasets are conducted by comparing the versions of integrating EM to NSGA-II, SPEA2 and MOEA/D with the original algorithms without EM. The results demonstrate that EM can improve the performance of NSGA-II and SPEA2, especially for NSGA-II.

Simulation models are frequently analyzed through a linear regression model that relates the input/output data behavior. However, in several situations, it happens that different data subsets may resemble different models. The purpose of this paper is to present a procedure for constructing switching regression metamodels in stochastic simulation, and to exemplify the practical use of statistical techniques of switching regression in the analysis of simulation results. The metamodel estimation is made using a mixture weighted least squares and the maximum likelihood method. The consistency and the asymptotic normality of the maximum likelihood estimator are establish. The proposed methods are applied in the construction of a switching regression metamodel. This paper gives special emphasis on the usefulness of constructing switching metamodels in simulation analysis.

This paper proposes a time-computing model using the Graphical Evaluation and Review Technique (GERT) to analyse concurrent New Product Development (NPD) processes. The research presented here differs from previous work carried out on concurrent engineering. First, we conceptualise a concurrent NPD process using the GERT scheduling technique and derive a method of modelling the information and communication complexities within the process. Second, we extend previous research carried out on concurrent engineering and incorporate it within our model. Finally, we present an alternative method of analysing concurrent NPD process for both researchers and project managers alike. The GERT model developed in this paper was successfully employed at two NPD firms located in Ireland and Iran.

Operational research assumes that organizational decision-making processes can be improved by making them more rigorous and analytical through the application of quantitative and qualitative modeling. However, we have only a limited understanding of how modeling actually affects organizational decision-making behavior, positively or negatively. Drawing from the Carnegie School's tradition of organizational research, this paper identifies two types of organizational decision-making activities where modeling can be applied: routine decision making and problem solving. These two types of decision-making activities have very different implications for model-based decision support, both in terms of the positive and negative behavioral impacts associated with modeling as well as the criteria used to evaluate models and modeling practices. Overall, the paper offers novel insights that help understand why modeling activities are successful (or not), explains why practitioners adopt some approaches more readily than others and points to new opportunities for empirical research and method development.

Group model building (GMB) is a participatory approach to using system dynamics in group decision-making and problem structuring. This paper considers the published quantitative evidence base for GMB since the earlier literature review by Rouwette et al. (2002), to consider the level of understanding on three basic questions: what does it achieve, when should it be applied, and how should it be applied or improved? There have now been at least 45 such studies since 1987, utilising controlled experiments, field experiments, pretest/posttest, and observational research designs. There is evidence of GMB achieving a range of outcomes, particularly with regard to the behaviour of participants and their learning through the process. There is some evidence that GMB is more effective at supporting communication and consensus than traditional facilitation, however GMB has not been compared to other problem structuring methods. GMB has been successfully applied in a range of contexts, but there is little evidence on which to select between different GMB tools, or to understand when certain tools may be more appropriate. There is improving evidence on how GMB works, but this has not yet been translated into changing practice. Overall the evidence base for GMB has continued to improve, supporting its use for improving communication and agreement between participants in group decision processes. This paper argues that future research in group model building would benefit from three main shifts: from single cases to multiple cases; from controlled settings to applied settings; and by augmenting survey results with more objective measures.

Manufacturers in the western world need to exploit and perfect all their strengths to reduce the flight of manufacturing to global outsourcing destinations. Use of automated manufacturing systems (AMSs) is one such strength that needs to be improved to perfection. One area for improvement is the management of uncertainties on the production floor. This paper explores strategies for modifying detailed event list schedules following the occurrence of an interruption. Advanced planning and scheduling (APS) software packages provide a detailed advance plan of production events. However, the execution of this advance plan is disrupted by a myriad of unanticipated interruptions, such as machine breakdowns, yield variations, and hot jobs. The alternatives available to respond to such interruptions can be classified in four groups: regenerating the complete schedule using APS, switching to dispatching mode, modifying the existing schedule, and continuing to follow the schedule and letting the production system gradually absorb the impact of the interruption. Regeneration of the complete schedule using APS requires a large computation effort, may result in large changes in the schedule, and hence is not recommended. This paper reports on an experimental study for evaluating 10 strategies for responding to machine failures in AMSs that broadly fall in the latter three groups. The strategies are evaluated using simulation under an experimental design with manufacturing scenario, load level, severity and duration of interruptions as factors. The results are analyzed to understand the strengths and weaknesses of the considered strategies and develop recommendations.

This paper argues that if OR is to prosper it needs to more closely reflect the needs of organizations and its practitioners. Past research has highlighted a gap between theoretical research developments, applications and the methods most frequently used in organizations. The scope of OR applications has also been contested with arguments as to the expanding boundaries of OR. But despite this, anecdotal evidence suggests that OR has become marginalized in many contexts. In order to understand these changes, IFORS (International Federation of OR Societies) in 2009 conducted a survey of global OR practice. The aim was to provide current evidence on the usage of OR tools, areas of application, and the barriers to OR’s uptake, as well as the educational background of OR practitioners. Results presented here show practitioners falling into three segments, which can be loosely characterized as those practicing ‘traditional’ OR, those adopting a range of softer techniques including Problem Structuring Methods (PSMs), and a Business Analytics cluster. When combined with other recent survey evidence, the use of PSMs and Business Analytics is apparently extending the scope of OR practice. In particular, the paper considers whether the Business Analytics movement, with an overlapping skill set to traditional OR but with a fast growing organizational base, offers a route to diminishing the gap between academic research and practice. The paper concludes with an exploration of whether this represents not just an opportunity for OR but also a serious challenge to its established practices.

Electrification of bus fleets in most cities is expected to rise due to its significant environmental benefits. However, electric buses have limited driving range and long recharging times. Additionally, electric buses require special charging infrastructure, which overall makes them less flexible than conventional diesel buses. Due to the limitations of the electric bus technologies, further adjustments have to be made to the current bus transport planning problems. The scheduling of electric vehicles is recognized as a fast-growing area of research. In this paper, we review 43 articles related to the electric bus technologies and give an overview of the different problems in the electric bus planning process (strategic, tactical and operational). The different problems are: 1) investment of electric bus fleet and charging infrastructure, 2) placement of charging infrastructure, 3) the electric vehicle scheduling problem (E-VSP) and 4) the charging scheduling problem. Given a set of timetabled trips and recharging stations, the E-VSP is concerned with finding a vehicle schedule that covers the trips and satisfies the driving range and recharging requirements of electric buses while minimizing operational cost. A detailed literature review of the constraints associated with the E-VSP and the solution approaches proposed to solve it is given. Rescheduling aspects or considerations of robustness for scheduling of electric vehicles is identified as a future area of research. Furthermore, integrated electric bus planning is considered as a crucial area of research and integrated approaches could further improve the efficiency of electric bus transport systems.

The wheels are one of the most worn components on a train. When the wear is unacceptable, the re-profiling can restore the shape of the wheel flange with the cost of decreasing the wheel diameter. The decision of re-profiling has serious implications for the life span of wheels. In this paper, based on the analysis of the wear and re-profiling characteristics of metro wheels, a data-driven model of the relationship between the wheel diameter, the flange thickness, their wear rates, and the re-profiling gain is built for the wheels of Guangzhou Metro Line One. A (SdP, SdR) re-profiling strategy is proposed, where SdP is the wheel flange thickness threshold to trigger a preventive re-profiling and SdR is the wheel flange thickness after the preventive re-profiling. Then the Monte Carlo simulation model of the re-profiling strategy is described in this paper. To find out when a re-profiling should be performed in terms of the flange wear-out level and what values of the flange thickness should be brought to by re-profiling, the simulation results for optimizing the decision variables (SdP, SdR) of the re-profiling strategy are given in this paper. Those having longer life spans are listed as the preferred re-profiling strategies. The study in this paper reveals that the wear rate of the flange thickness is correlated with the flange thickness, while the diameter wear rate could be considered independent of the flange thickness in terms of the wheels of Guangzhou Metro Line One. On the other hand, based on the observation and analysis of an available sample set from Guangzhou Metro Line One, the re-profiling gain is dependent on the flange thickness before or after re-profiling. The preferred re-profiling strategies suggested by this study can increase the life span comparing with the existing re-profiling strategies based on the simulation. The models and methods presented in this paper could benefit both city metro companies and inter-city rail companies by prolonging the life span of rolling stock wheels.

Every year, natural and man-made disasters affect hundreds of thousands of people and cause extensive damage. OR has made substantial contributions to disaster response and these have been the subject of several recent literature reviews. However, these reviews have also identified research gaps for OR – two of which are (1) limited contribution from soft OR, and (2) a need to model communications during disasters where there are complex interactions between stakeholders. At the intersection of these gaps we apply the Viable System Model (VSM) to examine challenges of rapid communication viability during dynamic disasters. The data that informs this paper were collected in four case studies in Japan – three on its current capabilities (e.g. a local government disaster management office) and one on its response to a past disaster (the Great Hanshin-Awaji Earthquake in 1995). This paper shows how applying VSM identified generic gaps and opportunities for communication systems and shows how these case studies signal the utility of VSM structures to arranging communications for fast-paced and changing environments. This paper also contributes to VSM theory through developing two new concepts (1) environmental support mechanisms for viability; and (2) rapid implementation unit emergence.

This paper presents a composite model in which two simulation approaches, discrete-event simulation (DES) and system dynamics (SD), are used together to address a major healthcare problem, the sexually transmitted infection Chlamydia. The paper continues an on-going discussion in the literature about the potential benefits of linking DES and SD. Previous researchers have argued that DES and SD are complementary approaches and many real-world problems would benefit from combining both methods. In this paper, a DES model of the hospital outpatient clinic which treats Chlamydia patients is combined with an SD model of the infection process in the community. These two models were developed in commercial software and linked in an automated fashion via an Excel interface. To our knowledge this is the first time such a composite model has been used in a healthcare setting. The model shows how the prevalence of Chlamydia at a community level affects (and is affected by) operational level decisions made in the hospital outpatient department. We discuss the additional benefits provided by the composite model over and above the benefits gained from the two individual models.

Additive manufacturing (AM) suggests promising manufacturing technologies, which complement traditional manufacturing in multiple areas, such as biomedical, aerospace, defense, and automotive industries. This paper addresses the production planning problem in multi-machine AM systems. We consider all relevant physical and technological parameters of the machines and the produced parts, for using direct metal laser sintering (DMLS) technology. In DMLS technology, each machine produces jobs, where each job consists of several parts arranged horizontally on the build tray. Starting a new job requires a setup operation. We address the simultaneous assignment of parts to jobs and jobs to the machines, while considering the cost and makespan objectives. A unified mixed-integer linear-programming (MILP) formulation that can minimize the above objectives separately and simultaneously is suggested, along with analytical bounds and valid inequalities. Experimentation demonstrates the effectiveness of the proposed formulation with single objectives versus similar formulations from the literature. An efficient frontier approach is applied to the multi-objective problem while generating a diverse set of exact non-dominated solutions. The trade-off between the objectives is analyzed via experimentation. Results show that when identical machines are used, the trade-off is relatively small, and hence the decision-maker can use any of the single objectives. However, when non-identical machines are used, it is important to consider both objectives simultaneously. Moreover, the trade-off increases with the number of machines and heterogeneity of the system, with respect to the size and settings of the machines.

In the context of big data, organizations and individuals can often benefit from the data mining techniques, such as classification. However, decision-makers must quickly react to insights over time under dynamic environments. In this paper, we present a novel perspective, named concept-cognitive computing system (C3S), to achieve dynamic classification learning over the partially labeled data and labeled data. More specifically, to store and consolidate knowledge, a concept falling space is first employed as a basic knowledge memory mechanism in C3S. Then, we design a new concept-cognitive process by means of simulating human learning processes, which can incorporate new information into the old knowledge. Finally, a strategy of constructing two different concept spaces is considered in our system when faced with the scenario of a partially labeled dynamic data. Although there exist significant differences between C3S and the conventional incremental learning methods in the learning paradigm, our proposed C3S still performs strong performance for dynamic classification in comparison with several state-of-the-art incremental learning approaches. In addition, the experiments on various datasets have demonstrated that our system can obtain a good performance on the partially labeled data and labeled data simultaneously in dynamic environments.

Itinerary planning serves an important purpose for travelers navigating through a public transit network. The uncertainty of travel times within the network must be considered since this uncertainty can lead to delays or missed transfers. If an initial transit service or a transfer between services is missed, then the traveler must have a contingency plan for reaching their destination. However, finding reliable itineraries with contingency plans is computationally challenging. This paper outlines an a priori policy that maximizes the reliability, i.e. the probability of reaching the destination within a travel time budget, while considering the most reliable primary itinerary and backup itineraries in case a transfer is missed. We develop a network search algorithm that finds the most reliable itinerary with backups, which accurately accounts for the uncertainty of travel times and frequency of alternative options in the recommendation. We compare our approach to state-of-the-art contingency research, Dynamic Journeying Under Uncertainty by Hame and Hakula (2013), and show that it more accurately calculates the reliability, is more intuitive to the traveler, and has a faster runtime. Additionally, we highlight the differences in reliability across times of day, travel time budgets, and city regions for a real-world transit network in Gottingen, Germany. This approach offers travelers an easy-to-understand tool for planning reliable public transit travel itineraries.

In this paper, we focus on a two-level production and distribution planning problem in the field of supply chain management, and examine the situations where the leader does not fully know the manufacturing technologies of the follower. In such a situation, the parameters representing the manufacturing technologies cannot be explicitly used to formulate the follower’s production planning problem. To overcome this difficulty, we propose formulations that implicitly express manufacturing technologies by using the input-output data observed from the production activities of the follower, incorporating the idea of data envelopment analysis (DEA). Assuming that the follower has multiple production facilities, we consider two possibilities of the observable input-output data and formulate two corresponding production planning problems; firstly, that only the input-output data aggregated for all the production facilities can be observed collectively, and secondly, that the input-output data for each of the production facilities can be observed separately. To clarify the validity of these DEA approaches, we compare them with the conventional formulation with technological coefficients using a numerical example.