Operations Research Community Research Articles

Experimental Matching of Instances to Heuristics for Constraint Satisfaction Problems.

Constraint satisfaction problems are of special interest for the artificial intelligence and operations research community due to their many applications. Although heuristics involved in solving these problems have largely been studied in the past, little is known about the relation between instances and the respective performance of the heuristics used to solve them. This paper focuses on both the exploration of the instance space to identify relations between instances and good performing heuristics and how to use such relations to improve the search. Firstly, the document describes a methodology to explore the instance space of constraint satisfaction problems and evaluate the corresponding performance of six variable ordering heuristics for such instances in order to find regions on the instance space where some heuristics outperform the others. Analyzing such regions favors the understanding of how these heuristics work and contribute to their improvement. Secondly, we use the information gathered from the first stage to predict the most suitable heuristic to use according to the features of the instance currently being solved. This approach proved to be competitive when compared against the heuristics applied in isolation on both randomly generated and structured instances of constraint satisfaction problems.

Shared mobility systems

Shared mobility systems for bicycles and cars have grown in popularity in recent years and have attracted the attention of the operational research community. Researchers have investigated several problems arising at the strategic, tactical and operational levels. This survey paper classifies the relevant literature under five main headings: station location, fleet dimensioning, station inventory, rebalancing incentives, and vehicle repositioning. It closes with some open research questions.

Total-factor energy efficiency with congestion

Total-factor energy efficiency (TFEE) assessment has received increasing attention in both operations research and energy economics communities. Earlier TFEE studies implicitly assume that production activity lies in the economic area, which precludes the possibility that the production activity lies in the non-economic area and thus suffers from congestion. This paper contributes to develop TFEE index by taking into account congestion effect. It starts with the definition of congested production technology, based on which several data envelopment analysis models are proposed to construct TFEE index with congestion. The models for quantifying energy inefficiency caused by congestion effect are also developed. We apply the proposed index to evaluate the energy efficiency performance of Chinese industrial sectors at province level in 2010–2012. It is found that TFEE with congestion can yield useful insights about the choice of proper ways to achieve energy efficiency improvement. A comparison with the empirical results under congestion-free production technology indicates that ignoring congestion effect may lead to significantly different TFEE scores when congestion effect does exist.

Combining metaheuristics with mathematical programming, constraint programming and machine learning

During the last years, interest on hybrid metaheuristics has risen considerably in the field of optimization and machine learning. The best results found for many optimization problems in science and industry are obtained by hybrid optimization algorithms. Combinations of optimization tools such as metaheuristics, mathematical programming, constraint programming and machine learning, have provided very efficient optimization algorithms. Four different types of combinations are considered in this paper: (1) Combining metaheuristics with complementary metaheuristics. (2) Combining metaheuristics with exact methods from mathematical programming approaches which are mostly used in the operations research community. (3) Combining metaheuristics with constraint programming approaches developed in the artificial intelligence community. (4) Combining metaheuristics with machine learning and data mining techniques.

Interior-Point Based Online Stochastic Bin Packing

Bin packing is an algorithmic problem that arises in diverse applications such as remnant inventory systems, shipping logistics, and appointment scheduling. In its simplest variant, a sequence of T items (e.g., orders for raw material, packages for delivery) is revealed one at a time, and each item must be packed on arrival in an available bin (e.g., remnant pieces of raw material in inventory, shipping containers). The sizes of items are i.i.d. samples from an unknown distributions, but the sizes are known when the items arrive. The goal is to minimize the number of non­empty bins (equivalently waste, defined to be the total unused space in non­empty bins). This problem has been extensively studied in the Operations Research and Theoretical Computer Science communities, yet all existing heuristics either rely on learning the distribution or exhibit ​o(T) additive suboptimality compared to the optimal offline algorithm only for certain classes of distributions (those with sublinear optimal expected waste). In this paper, we propose a family of algorithms which are the first truly distribution-­oblivious algorithms for stochastic bin packing, and achieve ​O(√T) additive suboptimality for all item size distributions. Our algorithms are inspired by approximate interior­-point algorithms for convex optimization. In addition to regret guarantees for i.i.d. sequences, we also prove a family of novel regret bounds for general non­i.i.d. input sequences, including guarantees for locally adversarially perturbed i.i.d. sequences. To the best of our knowledge these are the first such results for non-­i.i.d. and non­-random-­permutation input sequences for online stochastic packing.

Guest Editorial: Special issue on constrained decision-making in robotics

Noname manuscript No. (will be inserted by the editor) Guest Editorial: Special Issue on Constrained Decision-Making in Robotics Marco Pavone · Stefano Carpin the date of receipt and acceptance should be inserted later As the complexity of tasks envisioned for robotic systems increases, their decision making module needs to be capable of concurrently evaluating and trading off multiple, possibly contradicting and stochastic ob- jectives. For example, a rescue robot might be required to plan trajectories so as to maximize the probability of success to reach a critical location and, at the same time, minimize the duration of the traversal. Other more general examples include trading off information gath- ering versus energy expenditures, computation time ver- sus optimality of a decision, or learning versus safety. A natural framework to address this class of problems is constrained decision-making, whereby a decision maker seeks to optimize a given cost function (often stochas- tic) while keeping other costs (usually involving risk assessments) below given bounds. In the last decade, the operations research community has made significant strides on the topics of constrained decision-making (notoriously more challenging than the unconstrained counterpart) and risk assessment in dynamic scenarios. The result is a comprehensive theory and a set of algo- rithmic tools for this class of decision-making problems. Yet, despite their relevance, these results have seen lim- ited application within the robotics domain. Accordingly, in July 2014 we organized the work- shop “Constrained decision-making in robotics: models, algorithms, and applications” during the Robotics Sci- ence and Systems (RSS) conference in Berkeley, CA. Marco Pavone Department of Aeronautics and Astronautics Stanford University E-mail: pavone@stanford.edu Stefano Carpin School of Engineering University of California, Merced E-mail: scarpin@ucmerced.edu The workshop had three objectives: (1) to convene to- gether researchers working in the areas of decision-making, risk theory, and robotics, (2) to inform robotic researchers about the state of the art in constrained decision-making and modern risk theory, and (3) to formulate a research agenda on the topics of risk modeling and constrained decision-making for robotic applications. After the workshop, participants were invited to sub- mit an extended version of their work. The call for con- tributions was also extended to the broader research community and widely circulated. Eventually, six pa- pers were selected for publication and are included in this special issue. Collectively, these papers covers sev- eral robotic application domains for constrained decision- making, ranging from sensors networks to autonomous vehicles and space robotics. The first three papers deal with a deterministic prob- lem setup. Specifically, the paper “Cognitive Robots Learning Failure Contexts through Real-World Exper- imentation” by Sertac Kapinar and Sanem Sariel stud- ies how to endow robots with the ability to learn about their limitations of performance when it comes to task execution, and to capitalize on such knowledge to for- mulate risk-averse decisions subject to performance con- straints. The authors tackle the problem by using an In- ductive Logic Programming (ILP) framework, whereby hypotheses are formulated that relate execution con- texts to action outcomes. Hypotheses are continuously updated based on online observations and are used to model risks, constraints, and limitations on task exe- cution. On top of this knowledge base, a planner for- mulates decision sequences fulfilling the constraints de- rived from the learned hypotheses. The theoretical frame- work is experimentally demonstrated and validated both on a mobile robot and on a 7-DOF robotic arm perform-

Operations Research challenges in forestry: 33 open problems

Forestry has contributed many problems to the Operations Research (OR) community. At the same time, OR has developed many models and solution methods for use in forestry. In this article, we describe the current status of research on the application of OR methods to forestry and a number of research challenges or open questions that we believe will be of interest to both researchers and practitioners. The areas covered include strategic, tactical and operational planning, fire management, conservation and the use of OR to address environmental concerns. The paper also considers more general methodological areas that are important to forestry including uncertainty, multiple objectives and hierarchical planning.

Some applications of polynomial optimization in operations research and real-time decision making

We demonstrate applications of algebraic techniques that optimize and certify polynomial inequalities to problems of interest in the operations research and transportation engineering communities. Three problems are considered: (1) wireless coverage of targeted geographical regions with guaranteed signal quality and minimum transmission power, (2) computing real-time certificates of collision avoidance for a simple model of an unmanned vehicle (UV) navigating through a cluttered environment, and (3) designing a nonlinear hovering controller for a quadrotor UV, which has recently been used for load transportation. On our smaller-scale applications, we apply the sum of squares (SOS) relaxation and solve the underlying problems with semidefinite programming. On the larger-scale or real-time applications, we use our recently introduced “SDSOS Optimization” techniques which result in second order cone programs. To the best of our knowledge, this is the first study of real-time applications of sum of squares techniques in optimization and control. No knowledge in dynamics and control is assumed from the reader.

Resilient Upgrade of Electrical Distribution Grids

Modern society is critically dependent on the services provided by engineered infrastructure networks. When natural disasters (e.g. Hurricane Sandy) occur, the ability of these networks to provide service is often degraded because of physical damage to network components. One of the most critical of these networks is the electrical distribution grid, with medium voltage circuits often suffering the most severe damage. However, well-placed upgrades to these distribution grids can greatly improve post-event network performance. We formulate an optimal electrical distribution grid design problem as a two-stage, stochastic mixed-integer program with damage scenarios from natural disasters modeled as a set of stochastic events. We develop and investigate the tractability of an exact and several heuristic algorithms based on decompositions that are hybrids of techniques developed by the AI and operations research communities. We provide computational evidence that these algorithms have significant benefits when compared with commercial, mixed-integer programming software.

Book review “Cell formation in industrial engineering: theory, algorithms and experiments” by Boris Goldengorin, Dmitry Krushinsky, Panos M. Pardalos

When I saw the title of the book, I thought of that as another sortie at bioinformatics by the operations research community. Indeed by googling “Cell formation” one gets within a lot of stuff on the biological cell processes. What a group of combinatorial optimization experts can do to advance a murky business of life developments? Yet by looking at the intro, I see the book is on “industrial cell formation”, which is a quite different business. It appears the cell formation problem (CFP) here belongs in a growing field of research related to the issues of partitioning a set of technology units in clusters that concentrate as many operations as possible within themselves so that those remaining in-between are minimized. I cannot help but bring inmymemories related to this “discovery”. That was exactly the problem posed before me by Arkady Gelman who unexpectedly entered my office in the Summer 1967 at which I, a fresh PhD in ComputationalMathematics, was trying to penetrate the mathematics of competitive equilibrium and fixed point theorems. This was the starting point of my lifelong addiction to cluster analysis research. Back then I formalized the Gelman’s problem as of finding a partition with the maximum within cluster summary similarities or, equivalently, the minimum between-cluster summary similarities. Jointly with my only collaborator of that time, late Volodia Kupershtokh (1945–2001), we came to the conclusion that the problem as is, at nonnegative similarities, has trivial solutions only. One way to go was to restrict ourselves with pre-specified cluster cardinalities at which we came up with what later was called Kernighan-Lin algorithm, and estimates that hold “almost always” ([1], Our first results: [2]). Another way tomove onwas foundwhenwe noticed that the problem

A study of constraint programming heuristics for the car-sequencing problem

In the car-sequencing problem, a number of cars have to be sequenced on an assembly line respecting several constraints. This problem was addressed by both Operations Research (OR) and Constraint Programming (CP) communities, either as a decision problem or as an optimization problem. In this paper, we consider the decision variant of the car sequencing problem and we propose a systematic way to classify heuristics for solving it. This classification is based on a set of four criteria, and we consider all relevant combinations for these criteria. Some combinations correspond to common heuristics used in the past, whereas many others are novel. Not surprisingly, our empirical evaluation confirms earlier findings that specific heuristics are very important for efficiently solving the car-sequencing problem (see for instance Smith, 1996), in fact, often as important or more than the propagation method. Moreover, through a criteria analysis, we are able to get several new insights into what makes a good heuristic for this problem. In particular, we show that the criterion used to select the most constrained option is critical, and the best choice is fairly reliably the “load” of an option. Similarly, branching on the type of vehicle is more efficient than branching on the use of an option. Overall, we can therefore indicate with a relatively high confidence which is the most robust strategy, or at least outline a small set of potentially best strategies.Last, following a remark in Régin and Puget (1997) stating that the notion of slack used in heuristics induces a pruning rule, we propose an algorithm for this method and experimentally evaluate it, showing that, although computationally cheap and easy to implement, this is in practice a very efficient way to solve car-sequencing benchmarks.

Applying structured scenario fusion to address multifaceted strategic challenges

This paper proposes an applied method by which to fuse multiple independently developed scenarios to serve as a basis for strategic studies in the national security operations research community. First, it describes the need to fuse scenarios methodically as a consequence of the evolution of global security conditions and as a matter of logic that might be otherwise overlooked. Second, it proposes applying Structured Scenario Fusion (SSF), a method that uses three stylized tables to systematically and transparently record the process of scenario fusion. Third, it provides an example application set in a hypothetical study of future security conditions in Somalia. It concludes that SSF offers a promising way to improve the pedigree, realism, and logic of assumptions underpinning scenario-based strategic studies, as well as the quality of strategic decisions and plans thereby justified.

Uncertain multiobjective traveling salesman problem

Traveling salesman problem is a fundamental combinatorial optimization model studied in the operations research community for nearly half a century, yet there is surprisingly little literature that addresses uncertainty and multiple objectives in it. A novel TSP variation, called uncertain multiobjective TSP (UMTSP) with uncertain variables on the arc, is proposed in this paper on the basis of uncertainty theory, and a new solution approach named uncertain approach is applied to obtain Pareto efficient route in UMTSP. Considering the uncertain and combinatorial nature of UMTSP, a new ABC algorithm inserted with reverse operator, crossover operator and mutation operator is designed to this problem, which outperforms other algorithms through the performance comparison on three benchmark TSPs. Finally, a new benchmark UMTSP case study is presented to illustrate the construction and solution of UMTSP, which shows that the optimal route in deterministic TSP can be a poor route in UMTSP.

Guest Editorial: Selected Papers of European Symposium of Algorithms

This Special Issue contains a selection of expanded, peer-reviewed papers originally presented at the 20th Annual European Symposium on Algorithms (ESA 2012) held in Ljubljana, Slovenia, on September 10–12, 2012. The ESA symposia are devoted to fostering and disseminating the results of high quality research on the design and evaluation of algorithms and data structures. The forum seeks original algorithmic contributions for problems with relevant theoretical and/or practical applications and aims at bringing together researchers in the computer science and operations research communities. The conference received 285 submissions from 40 countries and the PC selected 69 papers among which we decided to pick seven excellent papers composing this Special Issue. These journal papers, which have been rigorously reviewed according to the journal standards, cover a wide range of algorithmic topics, thereby appealing to both the experts in the field as well as to those who wish a snapshot of the current breadth of research in this area. The paper on “A Model for Minimizing Active Processor Time” considers an elementary scheduling problem over a collection of n jobs, each having an integer length and a set of time intervals in which it can be feasibly scheduled. Given a parameter B, the processor can schedule up to B jobs at a timeslot t so long as it is “active” at t . The goal is to schedule all the jobs in the fewest number of active timeslots. The machine consumes a fixed amount of energy per active timeslot, regardless of the number of jobs scheduled in that slot (as long as the number of jobs is non-zero). The authors present a linear-time algorithm for the case where jobs are unit length and have a

Multiobjective Travelling Salesman Problem under Stochastic Environment

Travelling salesman problem is a fundamental combinatorial optimization model studied in the operations research community, and yet, there is surprisingly little literature that addresses stochastic uncertainties and multiple objectives in it simultaneously. This paper is devoted to a novel TSP variation called stochastic multiobjective TSP (SMOTSP) with random variables on the arc, and a new solution approach is proposed to obtain Pareto efficient route in it, whose validity is proved finally.

Most Productive Scale Size versus Demand Fulfillment: A Solution to the Capacity Dilemma

The field of economics tends to associate capacity planning with economic scale size and find the characteristics of the production function whereas the field of operations research community tends to focus on demand fulfillment to reduce the loss of sales or inventory for profit maximization. However, it is troublesome capacity dilemmas for firms that need to achieve economic scale size and demand fulfillment simultaneously; in particular, firms operate in stochastic environments. This study fills a gap between these two. To improve capacity planning, this study proposes a multi-objective mathematical programming with data envelopment analysis (DEA) constraints. In particular, compromise programming sets a target which shows a tradeoff between the most-productive-scale-size (MPSS) benchmark and a potential demand fulfillment benchmark. In addition, the minimax regret (MMR) approach and the stochastic programming (SP) technique are used to address target variation caused by demand fluctuations. This study pushes the ex-post DEA analysis of efficiency estimation (i.e. position) towards the ex-ante DEA analysis of production planning (i.e. direction). The result shows that the proposed models provide managerial insights to address the capacity dilemma.

Policy analytics need more than a spreadsheet: a case study in funding formulae

This article presents two case studies, concerning the allocation of £Billions by a mechanism communicated via spreadsheet models. It argues that technical analytic skills as well as policy development skills are a vital component of governance. In the UK, Central Government uses funding formulae to distribute money to local service providers. One commonly stated goal of such formulae is equity of service provision. However, given the complexity of public services, together with variations in need, delivery style and the exercise of stakeholder judgement as to which needs should be met and how, such formulae frequently obscure the process by which equity has been taken into account. One policy ‘solution’ to managing such tensions is to seek ‘transparency’. With respect to funding formulae, this commonly involves publishing the underlying data and formulae in spreadsheets. This paper extends the argument that such ‘transparency’ requires an audience that understands the policy assumptions (and related conceptualisations), data sources, methodological approaches and interpretation of results. It demonstrates how the search for policy ‘transparency’ is also met by the technical quality assurance goals that the operational research community would recognise as best practice in the development both of software generally and spreadsheet models specifically. Illustrative examples of complex formulae acting to subvert equity are drawn from the English Fire and Rescue Service and Police Service allocation formulae. In the former, an increase in the amount of deprivation, as measured by one of six indicators, has the perverse effect of decreasing the financial allocation. In the latter, metropolitan areas such as London are found to gain most from the inclusion of variables measuring sparsity. The conclusion from these scenarios is that the steps needed to for technical quality assurance and policy transparency are mutually reinforcing goals, with policy analysts urged to make greater use of technical analytic skills in software development.

Combining metaheuristics with mathematical programming, constraint programming and machine learning

During the last years, interest on hybrid metaheuristics has risen considerably in the field of optimization and machine learning. The best results found for many optimization problems in science and industry are obtained by hybrid optimization algorithms. Combinations of optimization tools such as metaheuristics, mathematical programming, constraint programming and machine learning, have provided very efficient optimization algorithms. Four different types of combinations are considered in this paper: (1) Combining metaheuristics with complementary metaheuristics. (2) Combining metaheuristics with exact methods from mathematical programming approaches which are mostly used in the operations research community. (3) Combining metaheuristics with constraint programming approaches developed in the artificial intelligence community. (4) Combining metaheuristics with machine learning and data mining techniques.

In This Issue

In This Issue

Data Mining for Industrial Engineering and Management

The focus areas of the Industrial Engineering and Management journal include production, logistics, quality, operational research, information systems, technology, communication, industrial economics, regional development, management, organizational behavior, human resources, finance, accounting, marketing, education, training, and professional skills [1]. The aim of this journal is to become a reliable source of information for leaders in the field of industrial engineering management journals research, and to feature a rapid review process [2]. The subject discussed in this paper is data mining for industrial engineering and management. Knowledge Discovery in Databases Knowledge Discovery in Databases (KDD) is an iterative process of extracting implicit, previously unknown, and potentially useful knowledge as a production factor from large datasets [3]. It includes data selection, cleaning, integration, transformation, data mining (DM), and reporting. The KDD process consists of steps that are performed before conducting data mining (i.e., selection, pre-processing, and transformation of data), the actual DM, and subsequent steps (i.e., interpretation, and evaluation of results) [4]. DM refers to the specific step of applying one or more statistical, machine-learning, or imageprocessing algorithms to a particular dataset with the intent to extract useful patterns from the datasets [5]. DM is widely used in market segmentation, customer profiling, fraud detection, retail promotions, and credit risk analysis [6]. Data Mining With the rapid growth of databases in numerous modern enterprises, DM has become an increasingly valuable data analysis approach. The operations research community has made substantial contributions to this field, particularly by formulating and solving numerous DM problems as optimization problems. In addition, several operations research applications can be addressed using DM methods [7]. In recent years, the data mining field has experienced substantial interest from both academia and industry. DM problems are typically categorized as association, clustering, classification, and prediction [8]. DM involves various techniques, including statistics, neural networks, decision trees, genetic algorithms, and visualization techniques that have been developed over the years. Statistics Regression is one of the most crucial statistical methods applied to science, engineering, economics, and management [9]. Regression is useful for the prediction of the presence or absence of a characteristic or outcome based on values of a set of independent variables that are continuous, categorical, or both. Furthermore, it assumes that measures of dependent variables were independently and randomly sampled, all potentially relevant independent variables are in the model, and all independent variables in the model are relevant [10,11]. Artificial neural networks