Book Reviews

Poor Economics

Banerjee, Abhijit, Esther Duflo. 2012. Poor Economics. Penguin Books. 303 pp. $15.99.

Poor Economics, a remarkable book about the poorest of the poor—those living on less than a dollar a day—is written by two MIT-based economists, using what operations research (OR) people might justifiably call the OR approach. I recommend it to all OR students who are interested in finding effective ways to help the poor.

Within a year of its publication, it has had rave reviews and plaudits from eminent writers on poverty, and has received the Financial Times and Goldman Sachs business book of the year award. Like all good OR work, it provides great insights into the problem it tackles, that is, effective ways to aid the poor; however, it also suffers from the weaknesses of the OR approach by underplaying the political dimensions of the problem.

The book’s accompanying website (http://www.pooreconomics.com) has a more enlightening title for the book: Poor Economics: A Radical Rethinking of the Way to Fight Global Poverty. The authors explain this title in the book’s foreword: “Poor Economics is ultimately about what the lives and choices of the poor tell us about how to fight global poverty” (p. xi).

The book is about effective policy interventions for helping the poor. It is the culmination of the authors’ work with scores of NGOs and private and public organizations in dozens of countries over more than 15 years. It starts by eschewing the popular but futile theoretical debate about whether aid is good or bad for the poor. Is Sachs (2005) right in claiming that if the world’s rich commit $195 billion in aid until 2025, poverty could be entirely eliminated by 2026? Or, does aid do more harm than good as Easterly (2001, 2006) and Moyo (2001) argue? Their answers are OR-like: instead of trying to find universal answers, let us look at concrete instances of what aid tries to achieve and bring evidence to bear on how we can better achieve it. Evidence includes the great mass of data now available, randomized controlled trials (RCTs) carried out by scores of researchers, and the traditional anecdotal evidence and expert opinions.

For example, the authors look at the problem of distributing insecticide-treated bed nets for malaria prevention, discuss various strategies based on the experiences of people working on the ground and of RCTs, and evaluate whether giving these nets for free or charging a small amount for them is better. Similarly, for immunization programs that have poor completion rates, even when vaccinations are free, they try different strategies, such as publicity blitzes, direct visits to the villages, and various economic incentives. They find dramatic improvements when lentils are given with vaccines.

This book includes many other impressive health-related examples, including cheap and effective ways to sanitize water, dealing with diarrhea, and providing breast feeding information. In each case, the reader learns about the ineffectiveness of old policies, and possible causes, cures, and RTC results that can lead to better outcomes. The authors use the same methodology on problems related to education, agricultural practices, and family planning. The book can teach the reader a lot about how the poor make their decisions and which interventions might work. It offers proof of the effectiveness of selected approaches by way of ex-post evaluations. The authors make readers feel that they are in the hands of masters of their craft.

In a glowing review, Amartya Sen, the Nobel Prize-winning economist, declares: “Banerjee and Duflo offer a new understanding of the surprising way the world really works” (quoted on the back cover). In my opinion, this is not quite true. Although the book gives great insights into how the poor make rational decisions, it fails to recognize that addressing poverty without engaging in a political struggle is sometimes impossible.

The fight against slavery had to be won before the lot of the slaves, the poorest of the poor in those times, could be improved. Under South Africa’s apartheid regime, dealing with the poverty within the majority black population without defeating the apartheid regime was impossible. In many parts of India, powerful upper-caste landlords subjugate the low-caste poor in rural areas; not surprisingly, the poor support the most extreme Maoist parties because these parties are the only ones prepared to take on the powerful.

To give them their due, the authors are aware of the political dimension. In the book’s last chapter, they talk about the importance of politics in developing and implementing good policies, but they confine themselves to looking only at good institution building for improvement at the margins. Examples include publishing information about school budgets, using mystery shoppers to monitor the efficiency of police stations, improving the design of ballot papers, and introducing quotas for female leadership of local government councils. Such policies are great and insightful, but not always applicable.

We can debate whether eliminating or reducing the power imbalance between the world’s rich and powerful and the rest of the world would be an effective way to help the poor. But, if you want to do something effective now under the prevailing circumstances, this is your book.

Gautam Appa

London School of Economics, London, United Kingdom, [email protected]

Integration of Information and Optimization Models for Routing in City Logistics

Ehmke, Jan Fabian. 2012. Integration of Information and Optimization Models for Routing in City Logistics. Springer. 213 pp. $129.00.

In Integration of Information and Optimization Models for Routing in City Logistics, Ehmke develops an approach to improve the routing of delivery vehicles within inner cities. He starts by including dynamic data to realistically reflect traffic conditions. His goal is to improve service quality by reducing the discrepancy between planned and actual data. In the first part of the book (Chapters 2 and 3), he presents the basics of city logistics, the concepts of attended home delivery, and the problems related to transport logistics. However, he does not always make clear the major differences between delivering to commercial customers and to end users.

In the second part (Chapters 4 and 5), the author points out the theoretical fundamentals of data analysis in this area, and uses it as a basis for processing the traffic data detected. He presents data mining techniques by focusing on cluster analysis and exploratory data analysis, and then describes approaches for data detection and collection. He begins this part by discussing the traditional approaches, which are largely based on stationary systems. Traffic data flow models are derived from this data, making suitable forecasts possible. He next presents telematics-based concepts, which are based on floating car data (FCD) approaches. To locate vehicles, a global positioning system (GPS) is applied and geographic information system (GIS) applications are used to make the data manageable and suitable for processing. Finally, he discusses how to use these approaches to process the data.

Chapters 6 and 7, the third part, address the integration of information models. They first address the creation of distance matrices, starting with static routing information based on digital roadmaps. However, these roadmaps are only partially suitable for providing information that is flexible, but time-dependent data structures can be developed based on this information. He uses data that depend on the daily load curve, thus enabling a detailed presentation of the traffic situation with respect to spatial distribution and the distribution of the traffic flow throughout the day.

In view of the complexity and the scope of the recoded data, these data are generated on a case-by-case basis and with the help of mechanisms for determining time-dependent shortest paths. The quality of the calculated data is evaluated in the next step, which uses various scenarios as a basis. The starting point is FCD-based information for the Stuttgart area. A comparison shows that the calculated data sufficiently match the dynamically recorded data. This provides an adequate basis for generating the necessary data to improve the routing of delivery processes. However, time-dependent data only reflect a momentary (static) structure because real-time information about the actual traffic situation is not available.

In the fourth and final part (Chapters 8 and 9), the author examines the application of optimization methods to provide solutions to various routing problems. First, he looks at the (uncapacitated) routing of one vehicle in detail, and applies the solution to the problem of the traveling salesman with static and time-dependent data. In the next step, he considers the routing of a fleet of vehicles by also taking time-dependent data structures as bases. Subsequently, he considers hard and soft time windows, which are important because customers request them. He shows that problems of this kind can be solved by using metaheuristics. He examines a single-vehicle and a vehicle-fleet problem under three spatial-differentiated scenarios from the Stuttgart area, and proves that time-dependent information has a significant impact on routing. Finally, it explores the effects of time windows within the scope of a simulation approach because of their importance to the routing results.

This book is interesting from both a theoretical and practical point of view. The author successfully connects the approaches of data mining and operations research to demonstrate possible ways to efficiently design delivery routes in inner-city areas by including time-dependent data. He shows and tests appropriate solutions. However, he does not include real-time information to allow efficient dispatching and online routing. As he notes, this approach could be an area of interest for additional research.

Joachim R. Daduna

Berlin School of Economics and Law, Berlin, Germany, [email protected]

Complementarity Modeling in Energy Markets

Gabriel, Steven A., Antonio J. Conejo, J. David Fuller, Benjamin F. Hobbs, Carlos Ruiz. 2012. Complementarity Modeling in Energy Markets. Springer. 655 pp. $119.00.

Mathematical (optimization) models are increasingly necessary for decision support in energy markets. This has led to a significant increase in the number of quantitative models in use, especially since the two recent oil crises and the liberalization of energy markets. Most large-scale models used up to now are based on traditional perfect-competition and deterministic approaches. One reason for their widespread use is the ease with which they can be used to solve real-world problems, especially linear problems with several million variables and constraints.

As a result of liberalizing and restructuring energy markets and progress made on using complementarity formulations (and other variations) to solve market equilibrium models, such approaches are becoming more important for modeling energy (and other) markets. The authors, three from North America and two from Europe, are well-known specialists in energy market modeling. Their objective in this book is to address the topic of complementarity modeling to make these modeling approaches widely accessible to a larger audience. Their approaches allow the modeling of noncooperative games in which each market agent can solve a separate but related optimization problem, subject to system-wide constraints. Additionally, this model formulation allows the inclusion of constraints on both primal and dual variables, which traditional problems cannot handle. The authors believe that such “improved models and algorithms in this area will therefore be fruitful and challenging areas of research” (p. ix) and that the “book serves as a useful guide in that direction” (p. ix). These statements aptly describe the book’s contribution. It fills a gap in this field of modeling and provides good input to the research community.

What prerequisites are necessary? I would not recommend this book for those who do not have a background in traditional modeling, particularly a basic knowledge of operations research (OR). The reader should understand topics such as traditional optimization models, duality, and Karush-Kuhn-Tucker conditions, although the book’s introduction includes a review of these topics to give readers a common knowledge level.

Who might be interested in such a book? I strongly recommend it for energy practitioners and researchers in energy market modeling, who are looking for methods to represent noncooperative games and hierarchical problems that result from microeconomic principles (i.e., bilevel models that anticipate market outcomes) and the interactions of market agents (i.e., equilibrium models).

What topics does the book address? After a short introduction that provides a reason for studying complementarity problems in the energy field, it introduces basic OR principles. This introduction addresses optimization and equilibrium problems, mathematical programs with equilibrium constraints (MPECs), equilibrium problems with equilibrium constraints (EPECs), and related topics, such as Karush-Kuhn-Tucker conditions, and sufficiency conditions for optimality, nonconvexity, and nonregularity issues. The introduction is useful in bringing readers to a level that allows them to differentiate the various classes of problems and related topics. Although this chapter is well written and includes several examples, a background in mathematical formulations and a basic knowledge of OR is necessary to understand both the introduction and the rest of the book. Chapter 3 explains necessary and useful principles of microeconomics. It starts with basics, such as supply and demand curves, and ends with Nash-Cournot models and the Stackelberg leader-follower model of imperfect competition. It builds the economic base that readers without an economics background need. Chapter 4 explores the notions of equilibria and optimization and their interrelationships, thus allowing the reader to more easily understand various characteristics of the systems under discussion. Subsequently, the authors introduce variational inequality problems, which sometimes have advantages over complementarity problems, as a natural alternative.

Chapters 6 and 7, the main part of the book, discuss bilevel problems, which explicitly indicate a hierarchy (i.e., hierarchical optimization). Chapter 6 addresses MPECs; Chapter 7 discusses EPECs. Examples in each chapter help the reader understand the concepts. Chapters 8 and 9 describe solution algorithms for complementarity problems, MPECs, and EPECs, techniques that are still being researched and developed. Chapters 10 and 11 discuss real-world modeling approaches for natural gas and electricity markets, respectively, including environmental issues. The final chapter, Chapter 12, deals with multicommodity markets, where linkages between markets, which are considered as equilibrium prices in one market, cannot be calculated without considering how they affect prices in other markets. The appendix includes several (but not all!) GAMS codes. This enhances the level of transparency and supports the reader in understanding the examples. Nevertheless, publishing all codes (e.g., in GAMS format on a website) would aid the reader’s comprehension and enhance the quality of the book.

Finally, traditional (large-scale) optimization problems based on simple methods (e.g., mixed-integer linear problems) will still be important in analyzing energy systems and markets in a competitive environment. The methods this book describe will not make these traditional approaches obsolete; however, I am convinced that they are a fruitful and challenging area of research, which will serve as a guide to real-world applications. Therefore, I strongly recommend this book to all modelers in energy systems and markets.

Dominik Möst

Technische Universität Dresden, Dresden, Germany, [email protected]

Community-Based Operations Research: Decision Modeling for Local Impact and Diverse Populations

Johnson, Michael P., ed. 2012. Community-Based Operations Research: Decision Modeling for Local Impact and Diverse Populations. Springer. 363 pp. $189.00.

There is just a chance that this book could open up a new field of activity for operations research (OR). And if it doesn’t, so much the worse for OR.

As I was reading the book, I had a continuous wrestling match with myself to understand what community-based operations research (CBOR) is, and what its boundaries and defining characteristics are. Looking back, I recall encountering the same difficulty with other new (to me) ideas. Some in the OR/MS community still have this type of cognitive immobility about problem-structuring methods, although these methods have been around and flourishing for three decades (at a conservative estimate). The novel element in CBOR, however, lies not in its transformative methodology, but in its application areas.

Back at the dawn of time (about 70 years ago), OR was only about the effectiveness of military operations. It then spread into areas such as heavy industry, commerce, and finance…but, Interfaces readers know all about that, so I will hurry past. Around the 1970s, model-based decision support was again spreading into nonmilitary government operations at the national and local levels (Quade 1975).

That is the context. So what is community-based OR? According to Michael Johnson, the volume’s editor, it is a subfield of public sector OR that emphasizes the needs and concerns of disadvantaged stakeholders in well-defined neighborhoods. It therefore overlaps with some types of decisions on which mainstream public sector analysts work and that have a necessarily local focus (e.g., decision making for emergency medical services). However, this book covers a largely novel range of topics: examples include managing community-based nonprofit organizations; sex offender management strategies; locating neighborhood parks; preventing children from exposure to lead; avoiding recruitment of vulnerable children to street gangs; and devising mass transit pricing policies that favor low-income riders. I speculate that a second edition or a new volume might also include managing food banks.

I think that this book will open the reader’s eyes to a range of possibilities for analytic work, which the management science and OR community has barely, if at all, touched on previously. Hopefully, the book will encourage some readers to explore the possibilities of engaging in such work, thereby helping to consolidate this new field.

What of the individual chapters? Almost all incorporate careful literature-survey sections on relevant preceding work, except where little of such work exists. The majority of the chapters use elaborate mathematical models, and sometimes multiple models, which they apply to local data. These contributions are addressed to a qualified professional audience, rather than to the disadvantaged people who are the center of concern.

One disappointment is that in many cases the models are used only on data collected from a neighborhood; that is, they were not actually used in hot blood to support decision makers in the relevant agencies or organizations. Doubtless, this is partly because of the path-breaking nature of the work. Potential sponsors of such work do not have the confidence boost that comes from the knowledge that there is a set of previously documented applications.

Interestingly, another relatively well-established branch of OR has a similar name and a related remit. That branch, community OR (COR), has been operating within the OR profession in Britain and more widely for the past 25 years (Midgley and Ochoa Arias 2004). These two branches share a common point of intellectual reference: the celebrated paper (Ackoff 1970) that discusses how Wharton School faculty worked with inhabitants of the local Mantua ghetto in Philadelphia. What then is the difference?

COR takes as its remit to work with (i.e., to take as its clients) disadvantaged community groups themselves. As a result, COR has less scope for mathematical models, which would at a stroke exclude this clientele from mental ownership of its own problem formulation. However, with appropriate low-tech analytic methods, it has proved possible to work closely with such groups, help them toward their goals, and even publish papers in academic journals about the engagements.

The opening chapters of this book, by Johnson and his collaborator Karen Smilowitz, provide a thorough and stimulating discussion of the nature of CBOR, and discuss what distinguishes it both from COR and from more conventional public sector OR. I hope that this promising initiative takes off; if so, this book will have a significant role in the development of our discipline. In the meantime, where else would you find an OR study of hair-care flow through salons in the black community?

Jonathan Rosenhead

London School of Economics, United Kingdom, [email protected]

Warehousing in the Global Supply Chain

Manzini, Riccardo, ed. 2012. Warehousing in the Global Supply Chain. Springer. 504 pp. $279.00.

Warehousing in the Global Supply Chain takes us through an almost chronological path in the development of warehouse operations and their positioning within the global supply chain. From the outset, the important role that warehouse functions play in the 21st century supply chain is recognized, independent of industry specifics, and the basic functions performed within a warehouse are observed. Various models and tools used in warehouse modeling are addressed and case studies to illustrate how to use different practices are included.

The book begins with a discussion of order picking, the most expensive and labor-intensive warehouse function. It describes order-picking strategies, performance issues, and the use of automation to support this functionality. I found the section on performance issues and measures helpful, particularly the images used to visualize the types of systems that have been engineered to suit various processes.

In Part I, we are walked through manual storage systems in a warehouse environment. The detail on warehouse layouts is succinct and the discussion of the optimal layout for storage and order-picking areas includes clear diagrams. A number of manual storage theories are discussed and include references to supporting literature. A concise summary of the theories is provided, clear definitions of the rules applied, and case studies included.

Part II addresses automated storage and retrieval systems (ASRS) and their application in material handling in distribution center and production environments. The information on facilities design hierarchy for a manufacturing plant describes the process at a high level, and includes support material on types and applications of ASRS. A broad range of high-level concepts is discussed; however, because the detailed breakdown on the associated systems is difficult to understand, a reader would need time to think about the material to understand it.

The glossary of measures included in the Designing Unit Load ASRS chapter would be helpful as a reference in designing a warehouse automation system or process. Although the material in this chapter is not easy to understand, I bookmarked it so that I can use the information in future studies. ASRS, a clearly mapped process model for designing unit load, could be helpful, even to a warehouse design novice, because it provides step-by-step instructions. The material on developing the objective function and the assumptions made are thought provoking. Warehouse Management: Productivity Improvement in Automated Storage and Retrieval Systems is a visually strong chapter, and is an appropriate introduction for the algorithms used in calculating travel times between warehouse storage locations.

Once a problem statement has been defined, the analytical and numerical modeling of cycle time within an ASRS can be used to determine the system requirements. In explaining the modeling process, this book describes how the models make it possible to predict the behavior of an ASRS, subject to defined operating parameters, as a way to ascertain performance.

One chapter, which focuses on new technology for unit load automated storage systems, describes the evolution of warehouse operations to facilitate the use of autonomous vehicle storage and retrieval systems and supporting models. Although it describes a specific solution, I found it interesting because of its application to heavy traffic conditions.

The chapter on Intelligent Optimization Methods for Industrial Storage Systems resonated with me. The logic sequences it describes are insightful, particularly those on swarm intelligence and ant/bee colony optimization, and reflect the next steps in systems and process optimization. By incorporating artificial intelligence, we can expect that the proposed systems will adapt to changes in the economic-technical world, and will allow more efficient use of resources, more economical and user-friendly production, and greater levels of independence through general evolutionary computation. On reflection, I see potential within a highly standardized environment; however, I also look to the future to understand how to better accommodate unexpected events in the next stages in the development of associated systems and methods.

Part III uses case studies to demonstrate how to apply the concepts explained earlier in the book. The first example is a case study from the tile industry. This chapter provides a basic approach for the storage allocation process to reduce travel time and distance by introducing storage and positioning rules. The next chapter looks at the design and optimization of order picking and how grouping orders can provide opportunities by optimizing travel distance. The final case study presents the application of the logistics reengineering process in a warehouse and order-fulfillment system of an Italian distribution company; this company uses warehouse management systems to improve time delivery rates, value-added productivity, and asset profitability, thus curtailing order-fulfillment lead times, and inventory and logistics costs.

The closing chapter, Warehouse Assessment in a Single Tour is well thought out; it considers multiple factors and poses questions related to throughput time, handling time, and process-improvement opportunities. The output of any assessment offers the potential for short- and long-term benefits through critical assessment of processes and development of justifications for investing in more sophisticated systems. One key point relates to using information technology to manage warehouse operations, a practice prevalent in 21st century supply chains. Numerous optimization opportunities are available for further development and refinement.

This book’s structure is academic, similar to a textbook. The technical language is sometimes difficult to follow; however, the range of the theories presented is broad. I like its chronological approach, in which the authors explain the history of warehouse planning and management and its development from manual to automated systems.

Directly applying the theories described by using the proposed models, techniques, and algorithms could be difficult; however, the book provides an adequate introduction to enable the reader to further research the concepts discussed.

The editor describes the book as a valuable source of information for engineering, doctoral, and postdoctoral students, and researchers in academic institutions. I agree strongly with this and recommend the book as a reference to this audience.

Bernard Stanley

TNT Express, Liverpool, United Kingdom, [email protected]

Mathematical Optimization of Water Networks

Martin, Alexander, Kathrin Klamroth, Jens Lang, Gunter Leugering, Antonio Morsi, Martin Oberlack, Manfred Ostrowski, Roland Rosen. 2012. Mathematical Optimization of Water Networks. Springer Basel. 210 pp. $109.00.

Mathematical Optimization of Water Networks presents new approaches to the simulation and optimization techniques used in water supply and sewage systems. It covers three areas: water networks (Chapters 1–4), wastewater using a shallow water approach (Chapters 6 and 8), and sewage network optimization (Chapters 5, 7, 9, and 10).

The book’s discussion of water networks addresses the complexity of pressurized flow through pipe networks, which consist of pipes, storage tanks, pumps, and valves. A system of differential algebraic equations generated in the process is tackled using the Rosenbrock-Wanner method, and the Lax-Friedrich approach and a weighted essentially nonoscillatory reconstruction are used for space discretization. The use of a unified modeling approach to cover free surface flow and pressured flow is a significant contribution. Next, a simulation tool is presented, which uses an implicit box scheme for the discretization of the water hammer equations to numerically solve the underlying model equations and the optimization problem, which is based on a gradient-based optimization technique. The derivation of adjoint calculus for water networks to efficiently solve the optimization problem is also useful. An implicit box scheme is used to effectively discretize the pipe dynamics represented by the governing partial differential water hammer equations. The nonlinearities of the discretized pipe and pump equations are addressed by using a piecewise linearization scheme. The implicit Euler discretization is successfully applied to discretize the ordinary differential equation of a tank’s filling level over time, and linearize the nonlinear discretized terms using piecewise linearization. Finally, a continuous nonlinear optimization technique is devised to address approximation errors that arise when developing the mixed-integer optimization model. The heuristic approach used combines and takes advantage of the mixed-integer optimization technique and continuous nonlinear optimization model (both introduced in Chapter 3) to find feasible solutions. The authors effectively present their approach and the computational results of a sample water supply network. However, they focus exclusively on explaining their models; I would find a comparison of their approach to existing models also interesting.

In applying shallow water theory, the authors first use the discontinuous galerkin (DG) finite-element method to address the limitation of Saint Venant’s equation, which is only valid for partially filled cross sections in a sewer network. The DG scheme solves the governing equations both in the free surface flow and pressurized pipe flow, and in the transition between the free surface flow and pressurized flow. Theoretical and numerical methods are discussed for computing the transition point (from free surface flow to pressurized flow) and free surface and pressurized flow regions. Sample problems are used to test the performance of the DG code, and then discrete adjoint computations based on Runge-Kutta schemes are applied to generate robust gradients for descent methods, and subsequently descent methods are used to generate an optimal control for the network. The application of a hydrodynamic process based on shallow water equations is a useful research technique in urban drainage systems control. The shallow water equations cannot account for the flow of the liquid across structures such as pumps, weirs, and storage containers. Therefore, the effects of these structures are incorporated into the equations. However, in the derivations, many assumptions are made, which challenge the optimality of the results. The boundary conditions used in the finite-volume method are not explained in detail.

In addressing sewage network optimization, an important and timely topic, first the strengths and weaknesses of the available optimization and simulation software are described, in particular, model predictive control (MPC), for optimal control of sewer networks; however, the MPC’s main computational and implementation challenges are not described in sufficient detail. Second, components in the MPC dynamic-process-model approach for solving the optimal sewer network problem are discussed, including SWWM 5 for a process model, BlueM.Opt for optimization, and additional software to control the receding-horizon process and supply information for the data flow. Next, two main simulation and optimization software packages are compared for the optimal sewer network with a dynamic-model problem. Previous research was purely academic work to assess the optimality of proposed software; the authors of this book focus on applying this software to realistic problems to ascertain their efficacy. I commend how they applied the software to a case study of a network in a small German city. Finally, they provide multiobjective analyses of wastewater to assist decision makers in making informed decisions based on trade-off data. They introduce three scalarization methods, which provide several solutions to the waste management problem, because decision making is usually informed by the alternative solutions available. The five main objective functions are optimized sequentially and scalarized later to provide one main objective function value, although the additional problem constraint introduced by the scalarization needs to be explained in detail. In future work, some stochastic concepts should be introduced into the optimization and scalarization, because the global optimality of the solution is not guaranteed. One disappointment in these chapters is that they cite insufficient references.

In summary, the authors recognize the complexity of a water supply and sewage system, and they develop mathematical models of the system to illustrate the theory underlying it and the trade-offs and compromises of its components; thus, they help with the decision side of the planning problem. They have made significant contributions to the modeling and optimization of water supply and sewage systems. Although this is not a comprehensive book covering all topics related to water supply network and sewage systems, it is an important reference to help academia and professionals appreciate new approaches in simulation and optimization techniques. My only negative comment is that I wish the authors had compared their approach to the methods discussed in the literature.

Shoou-Yuh Chang

Department of Civil Engineering, NC A&T State University, Greensboro, North Carolina 27411, [email protected]

Advanced Planning in Supply Chains

Stadtler, Hartmut, Bernhard Fleischmann, Martin Grunow, Herbert Meyr, Christopher Sürie. 2012. Advanced Planning in Supply Chains. Springer. 313 pp. $79.95.

In Advanced Planning in Supply Chains, the authors present a unifying framework for illustrating the planning tasks an organization must perform, and discuss how an advanced planning system (APS) could help the organization cope with planning complexity. However, an APS can be complex and hard to understand, as is illustrated by the statement that “Listening to a Stradivari played by an expert is a real treat, but if it is played by a beginner…” (p. v). As a practitioner, I wholeheartedly agree. It shows that the authors understand the concepts and benefits, but also the many pitfalls, of an APS in practice.

The book is built around Frutado, a fictitious company in the consumer goods industry. The authors describe the company’s planning tasks in detail, explain supply chain concepts, and provide possible models to solve. They use the SAP advanced planning and optimization software (SAP APO) to show implementation and customization aspects. SAP APO includes interactive learning units that a user can download. Using the system in either of its modes, demo or practice, requires only an Internet browser. Demo mode shows what a user must do to set up a particular planning task; practice mode enables the user to do some exercises without having access to a full SAP APO system.

This book has three parts: Part I describes Frutado, including its supply chain and the planning tasks at several levels of the chain. The supply chain planning matrix and hierarchical planning are used to decompose and explain how the planning task models could be solved. The approach used is practical; the authors make clear that one model cannot address all planning tasks. Part I also introduces SAP APO, including its architecture and the master data it requires; the introduction also explains how to embed SAP APO within a company’s information technology systems (from both a technical and a business perspective).

Part II, the main part of the book, has six chapters, each of which discusses a SAP APO module. Each chapter applies the same logic: First, it explains the conceptual background, referencing other textbooks and research papers. Then it gives a short introduction to Frutado’s specific learning units. The six chapters comprise:

• Demand Planning. This chapter discusses the basics of forecasting, the data needed, and the planning tasks at this level.

• Master Planning—Supply Network Planning. Based on the forecasts, medium-term plans can be developed. The chapter presents a basic linear programming model and some extensions using a mixed-integer programming model.

• Production Planning and Detailed Scheduling. For situations in which setup times and costs are important, this chapter explains the concepts of lot sizing and production orders. Simple heuristics and genetic algorithms are used to solve the scheduling task.

• Global Available to Promise. In global available-to-promise planning, the short-term customer orders are matched to supplies to maximize customer satisfaction. This chapter uses rule-based searching techniques to illustrate how to solve this planning issue.

• Deployment. The medium-term distribution planning task matches shortages and excesses in supply with the demand in the deployment from plants to customers. This chapter illustrates fair-share rules and push rules.

• Transportation Planning/Vehicle Scheduling. This chapter discusses the short-term planning of transportation processes, taking into account customer delivery dates and vehicle capacity constraints. SAP APO’s transportation and vehicle routing module is explained.

Finally, Part III discusses some APS implementation issues. This is a brief discussion (only five pages in a 300-page book) of some aspects that might arise before and during implementation. Although it is good as a high-level summary, it is no more than that.

The authors state that this book is a good introductory textbook for students. I would like to go further; I think it can also be valuable to both beginning and experienced supply chain (optimization) consultants, and also to organizational staff who work in the planning function. It can help these planners acquire a realistic view of how supply chain planning using an APS might work, and what a user might expect. In my opinion, this book can be instrumental in closing the gap between the theory and practice of supply chain planning.

Hein Fleuren

Department of Econometrics and Operations Research, Tilburg University, Tilburg, The Netherlands, [email protected]

Fundamentals of Queuing Systems: Statistical Methods for Analyzing Queuing Models

Thomopoulos, Nick T. 2012. Fundamentals of Queuing Systems: Statistical Methods for Analyzing Queuing Models. Springer. 170 pp. $109.00.

Fundamentals of Queuing Systems: Statistical Methods for Analyzing Queuing Models is a queuing theory book with practitioners as its target audience. This is obvious from the moment the reader opens it to find a plentiful supply of summary equations on each page; these equations describe numerous properties and results from relatively standard models in queueing theory. However, the consideration of a number of special cases permits the author to provide tailored solutions to specific problems, an ideal situation for the reader who is looking for some formulae for those problems.

Problems considered in the book range from queues with infinite streams of arrivals to finite arrival populations, and from single-service facilities to multiple-service facilities. They also span the usual range of exponential, deterministic, Erlang, and general service distributions.

Many real-world numerical examples are scattered throughout the book to provide the reader with opportunities to see how the formulae work. Many of the examples are numerical and tables dominate words, making them difficult to understand at first glance.

For the uninitiated, the book begins by providing rudimentary coverage of Poisson, exponential, and Erlang distributions, although a reader who is not familiar with these topics is probably better advised to first research them elsewhere. Following the Introduction and Preliminary Concepts chapters, each chapter addresses a particular queuing model; examples in the early chapters include M/M/1, M/M/1/N, M/M/1/1, M/M/k, and M/M/k/N. This continues throughout the 23 chapters, although models with repairmen (finite-customer populations), models with repeat service requirements, tandem queues, and problems with Erlang distributions are more typical of the later chapters. A number of early chapters provide the solutions to the basic queuing balance and equilibrium equations in special cases with one server: which are then covered again soon after in a more general multiserver setting. This is useful as a quick reference guide should one find oneself with a one-server problem; however, it is an approach that has obvious drawbacks by making the material appear repetitive.

In the interest of unwavering consistency, many chapters have similar layouts. The advantage is that information provided in a particular chapter is not given greater prominence than information in any other chapter. The reader can choose which aspects of a problem are most important and select only those results. Many subsection headings are uninformative and out of context with the material that follows, and some readers may become frustrated. Although the language the book uses is at times unpolished, it is comprehensible.

This book is not a textbook on queuing theory; it rarely offers insights on why results are true or tries to help the reader understand principles that connect different areas. An owner is unlikely to use it much; however, that owner may wish to occasionally browse it to see the standard kinds of results that exist in the area in which he (she) is interested. It is definitely a book to be browsed, rather than read linearly. Readers who are not interested in the theory underlying queuing theory results may find it a useful resource; however, the lack of contextualization and rigor in its arguments and explanations may make the reader wary of taking all given formulae as definitive. In conclusion, I do not consider this book an important addition to many bookshelves, except those of readers who know that they frequently have the need to reference its examples and models.

David Hodge

School of Mathematical Sciences, University of Nottingham, Nottingham, United Kingdom, [email protected]