 Open Access
 Total Downloads : 32
 Authors : A. Vigneshwari
 Paper ID : IJERTCONV5IS04008
 Volume & Issue : NCETCPM – 2017 (Volume 5 – Issue 04)
 Published (First Online): 24042018
 ISSN (Online) : 22780181
 Publisher Name : IJERT
 License: This work is licensed under a Creative Commons Attribution 4.0 International License
Operations Research
A. Vigneshwari
UG scholar, Imayam College of Engineering, Tamil Nadu, India
Abstract – This chapter will provide an overview of Operations Research (O.R.) from the perspective of an industrial engineer. The focus of the chapter is on the basic philosophy behind O.R. and the socalled "O.R. approach" to solving design and operational problems that industrial engineers commonly encounter. In its most basic form, O.R. may be viewed as a scientific approach to solving problems; it abstracts the essential elements of the problem into a model, which is then analyzed to yield an optimal solution for implementation. The mathematical details and the specific techniques used to build and analyze these models can be quite sophisticated and are addressed elsewhere in this handbook; the emphasis of this chapter is on the approach. A brief review of the historical origins of O.R. is followed by a detailed description of its methodology. The chapter concludes with some examples of successful realworld applications of O.R.
Keywords: Operation Research, optional problems, Scientific Approach, simplex method, PERT/CPM
1 INTRODUCTION
Although it is a distinct discipline in its own right, Operations Research (O.R.) has also become an integral part of the Industrial Engineering (I.E.) profession. This is hardly a matter of surprise when one considers that they both share many of the same objectives, techniques and application areas. O.R. as a formal subject is about fifty years old and its origins may be traced to the latter half of World War II. Most of the O.R. techniques that are commonly used today were developed over (approximately) the first twenty years following its inception. During the next thirty or so years the pace of development of fundamentally new O.R. methodologies has slowed somewhat. However, there has been a rapid expansion in (1) the breadth of problem areas to which
O.R. has been applied, and (2) in the magnitudes of the problems that can be addressed using O.R. methodologies. Today, operations research is a mature, welldeveloped field with a sophisticated array of techniques that are used routinely to solve problems in a wide range of application areas.
This chapter will provide an overview of O.R. from the perspective of an Industrial Engineer. A brief review of its historical origins is first provided. This is followed by a detailed discussion of the basic philosophy behind O.R. and the socalled "O.R. approach." The chapter concludes with several examples of successful applications to typical problems that might be faced by an Industrial Engineer. Broadly speaking, an O.R. project comprises three steps:

building a model, (2) solving it, and (3) implementing
the results. The emphasis of this chapter is on the first and third steps. The second step typically involves specific methodologies or techniques, which could be quite sophisticated and require significant mathematical development. Several important methods are overviewed elsewhere in this handbook. The reader who has an interest in learning more about these topics is referred to one of the many excellent texts on O.R. that are available today and that are listed under "Further Reading" at the end of this chapter, e.g., Hillier and Lieberman (1995), Taha (1997) or Winston (1994).

A HISTORICAL PERSPECTIVE
While there is no clear date that marks the birth of O.R., it is generally accepted that the field originated in England during World War II. The impetus for its origin was the development of radar defense systems for the Royal Air Force, and the first recorded use of the term Operations Research is attributed to a British Air Ministry official named A. P. Rowe who constituted teams to do "operational researches" on the communication system and the control room at a British radar station. The studies had to do with improving the operational efficiency of systems (an objective which is still one of the cornerstones of modern O.R.). This new approach of picking an "operational" system and conducting "research" on how to make it run more efficiently soon started to expand into other arenas of the war. Perhaps the most famous of the groups involved in this effort was the one led by a physicist named P. M. S. Blackett which included physiologists, mathematicians, astrophysicists, and even a surveyor. This multifunctional team focus of an operations research project group is one that has carried forward to this day. Blacketts biggest contribution was in convincing the authorities of the need for a scientific approach to manage complex operations, and indeed he is regarded in many circles as the original operations research analyst.
O.R. made its way to the United States a few years after it originated in England. Its first presence in the U.S. was through the U.S. Navys Mine Warfare Operations Research Group; this eventually expanded into the Antisubmarine Warfare Operations Research Group that was led by Phillip Morse, which later became known simply as the Operations Research Group. Like Blackett in Britain, Morse is widely regarded as the "father" of O.R. in the United States, and many of the distinguished scientists and mathematicians that he led went on after the end of the war to become the pioneers of O.R. in the United States.
In the years immediately following the end of World War II, O.R. grew rapidly as many scientists realized that the principles that they had applied to solve problems for the military were equally applicable to many problems in the civilian sector. These ranged from shortterm problems such as scheduling and inventory control to longterm problems such as strategic planning and resource allocation. George Dantzig, who in 1947 developed the simplex algorithm for Linear Programming (LP), provided the single most important impetus for this growth. To this day, LP remains one of the most widely used of all O.R. techniques and despite the relatively recent development of interior point methods as an alternative approach, the simplex algorithm (with numerous computational refinements) continues to be widely used. The second major impetus for the growth of O.R. was the rapid development of digital computers over the next three decades. The simplex method was implemented on a computer for the first time in 1950, and by 1960 such implementations could solve problems with about 1000 constraints. Today, implementations on powerful workstations can routinely solve problems with hundreds of thousands of variables and constraints. Moreover, the large volumes of data required for such problems can be stored and manipulated very efficiently.
Once the simplex method had been invented and used, the development of other methods followed at a rapid pace. The next twenty years witnessed the development of most of the O.R. techniques that are in use today including nonlinear, integer and dynamic programming, computer simulation, PERT/CPM, queuing theory, inventory models, game theory, and sequencing and scheduling algorithms. The scientists who developed these methods came from many fields, most notably mathematics, engineering and economics. It is interesting that the theoretical bases for many of these techniques had been known for years, e.g., the EOQ formula used with many inventory models was developed in 1915 by Harris, and many of the queuing formulae were developed by Erlang in 1917. However, the period from 1950 to 1970 was when these were formally unified into what is considered the standard toolkit for an operations research analyst and succesfully applied to problems of industrial significance. The following section describes the approach taken by operations research in order to solve problems and explores how all of these methodologies fit into the O.R. framework.

WHAT IS OPERATIONS RESEARCH?
A common misconception held by many is that
O.R. is a collection of mathematical tools. While it is true that it uses a variety of mathematical techniques, operations research has a much broader scope. It is in fact a systematic approach to solving problems, which uses one or more analytical tools in the process of analysis. Perhaps the single biggest problem with O.R. is its name; to a layperson, the term "operations research" does not conjure up any sort of meaningful image! This is an unfortunate consequence of the fact that the name that A. P. Rowe is credited with first assigning to the field was somehow
never altered to something that is more indicative of the things that O.R. actually does. Sometimes O.R. is referred to as Management Science (M.S.) in order to better reflect its role as a scientific approach to solving management problems, but it appears that this terminology is more popular with business professionals and people still quibble about the differences between O.R. and M.S. Compounding this issue is the fact that there is no clear consensus on a formal definition for O.R. For instance, C.
W. Churchman who is considered one of the pioneers of
O.R. defined it as the application of scientific methods, techniques and tools to problems involving the operations of a system so as to provide those in control of the system with optimum solutions to problems.

THE OPERATIONS RESEARCH APPROACH Given that O.R. represents an integrated
framework to help make decisions, it is important to have a clear understanding of this framework so that it can be applied to a generic problem. To achieve this, the so called O.R. approach is now detailed. This approach comprises the following seven sequential steps: (1) Orientation, (2) Problem Definition, (3) Data Collection,
(4) Model Formulation, (5) Solution, (6) Model Validation and Output Analysis, and (7) Implementation and Monitoring. Tying each of these steps together is a mechanism for continuous feedback; Figure 1 shows this schematically.
Figure 1: The Operations Research Approach
Each of these steps is now discussed in further detail. Orientation: The first step in the O.R. approach is referred to as problem orientation. The primary objective of this step is to constitute the team that will address the problem at hand and ensure that all its members have a clear picture of the relevant issues. It is worth noting that a distinguishing characteristic of any O.R. study is that it is done by a multifunctional team. To digress slightly, it is
also interesting that in recent years a great deal has been written and said about the benefits of project teams and that almost any industrial project today is conducted by multi functional teams. Even in engineering education, teamwork has become an essential ingredient of the material that is taught to students and almost all academic engineering programs require team projects of their students. The team approach of O.R. is thus a very natural and desirable phenomenon.
Problem Definition: This is the second, and in a significant number of cases, the most difficult step of the O.R. process. The objective here is to further refine the deliberations from the orientation phase to the point where there is a clear definition of the problem in terms of its scope and the results desired. This phase should not be confused with the previous one since it is much more focused and goal oriented; however, a clear orientation aids immeasurably in obtaining this focus. Most practicing industrial engineers can relate to this distinction and the difficulty in moving from general goals such "increasing productivity" or "reducing quality problems" to more specific, welldefined objectives that will aid in meeting these goals.
A clear definition of the problem has three broad components to it. The first is the statement of an unambiguous objective. Along with a specification of the objective it is also important to define its scope, i.e., to establish limits for the analysis to follow. While a complete system level solution is always desirable, this may often be unrealistic when the system is very large or complex and in many cases one must then focus on a portion of the system that can be effectively isolated and analyzed. In such instances it is important to keep in mind that the scope of the solutions derived will also be bounded. Some examples of appropriate objectives might be (1) "to maximize profits over the next quarter from the sales of our products," (2) "to minimize the average downtime at workcenter X," (3) "to minimize total production costs at Plant Y," or (4) "to minimize the average number of late shipments per month to customers."
The second component of problem definition is a specification of factors that will affect the objective. These must further be classified into alternative courses of action that are under the control of the decision maker and uncontrollable factors over which he or she has no control. For example, in a production environment, the planned production rates can be controlled but the actual market demand may be unpredictable (although it may be possible to scientifically forecast these with reasonable accuracy). The idea here is to form a comprehensive list of all the alternative actions that can be taken by the decision maker and that will then have an effect on the stated objective. Eventually, the O.R. approach will search for the particular course of action that optimizes the objective.
The third and final component of problem definition is a specification of the constraints on the courses of action, i.e., of setting boundaries for the specific actions that the
decisionmaker may take. As an example, in a production environment, the availability of resources may set limits on what levels of production can be achieved. This is one activity where the multifunctional team focus of O.R. is extremely useful since constraints generated by one functional area are often not obvious to people in others. In general, it is a good idea to start with a long list of all possible constraints and then narrow this down to the ones that clearly have an effect on the courses of action that can be selected. The aim is to be comprehensive yet parsimonious when specifying constraints.
Continuing with our hypothetical illustration, the objective might be to maximize profits from the sales of the two products. The alternative courses of action would be the quantities of each product to produce next month, and the alternatives might be constrained by the fact that the amounts of each of the three resources required to meet the planned production must not exceed the expected availability of these resources. An assumption that might be made here is that all of the units produced can be sold. Note that at this point the entire problem is stated in words; later on the O.R. approach will translate this into an analytical model.
Data Collection: In the third phase of the O.R. process data is collected with the objective of translating the problem defined in the second phase into a model that can then be objectively analyzed.
Model Formulation: This is the fourth phase of the O.R. process. It is also a phase that deserves a lot of attention since modeling is a defining characteristic of all operations research projects. The term "model" is misunderstood by many, and is therefore explained in some detail here. A model may be defined formally as a selective abstraction of reality. This definition implies that modeling is the process of capturing selected characteristics of a system or a process and then combining these into an abstract representation of the original. The main idea here is that it is usually far easier to analyze a simplified model than it is to analyze the original system, and as long as the model is a reasonably accurate representation, conclusions drawn from such an analysis may be validly extrapolated back to the original system.
Models may be broadly classified into four categories: Physical Models, Analogic Models, Computer Simulation Models, Mathematical Models
Mathematical Models: This is the final category of models, and the one that traditionally has been most commonly identified with O.R. In this type of model one captures the characteristics of a system or process through a set of mathematical relationships. Mathematical models can be deterministic or probabilistic. In the former type, all parameters used to describe the model are assumed to be known (or estimated with a high degree of certainty). With probabilistic models, the exact values for some of the parameters may be unknown but it is assumed that they are
capable of being characterized in some systematic fashion (e.g., through the use of a probability distribution). As an illustration, the Critical Path Method (CPM) and the Program Evaluation and Review Technique (PERT) are two very similar O.R. techniques used in the area of project planning. However, CPM is based on a deterministic mathematical model that assumes that the duration of each project activity is a known constant, while PERT is based on a probabilistic model that assumes that each activity duration is random but follows some specific probability distribution (typically, the Beta distribution). Very broadly speaking, deterministic models tend to be somewhat easier to analyze than probabilistic ones; however, this is not universally true.
Most mathematical models tend to be characterized by three main elements: decision variables, constraints and objective function(s). Decision variables are used to model specific actions that are under the control of the decision maker. An analysis of the model will seek specific values for these variables that are desirable from one or more perspectives. Very often especially in large models it is also common to define additional "convenience" variables for the purpose of simplifying the model or for making it clearer. Strictly speaking, such variables are not under the control of the decisionmaker, but they are also referred to as decision variables.Constraints are used to set limits on the range of values that each decision variable can take on, and each constraint is typically a translation of some specific restriction (e.g., the availability of some resource) or requirement (e.g., the need to meet contracted demand). Clearly, constraints dictate the values that can be feasibly assigned to the decision variables, i.e., the specific decisions on the system or process that can be taken. The third and final component of a mathematical model is the objective function. This is a mathematical statement of some measure of performance (such as cost, profit, time, revenue, utilization, etc.) and is expressed as a function of the decision variables for the model. It is usually desired either to maximize or to minimize the value of the objective function, depending on what it represents. Very often, one may simultaneously have more than one objective function to optimize (e.g., maximize profits and minimize changes in workforce levels, say). In such cases there are two options. First, one could focus on a single objective and relegate the others to a secondary status by moving them to the set of constraints and specifying some minimum or maximum desirable value for them. This tends to be the simpler option and the one most commonly adopted. The other option is to use a technique designed specifically for multiple objectives (such as goal programming).
In using a mathematical model the idea is to first capture all the crucial aspects of the system using the three elements just described, and to then optimize the objective function by choosing (from among all values for the decision variables that do not violate any of the constraints specified) the specific values that also yield the most desirable (maximum or minimum) value for the objective
function. This process is often called mathematical programming. Although many mathematical models tend to follow this form, it is certainly not a requirement; for example, a model may be constructed to simply define relationships between several variables and the decision maker may use these to study how one or more variables are affected by changes in the values of others. Decision trees, Markov chains and many queuing models could fall into this category.
Before concluding this section on model formulation, we return to our hypothetical example and translate the statements made in the problem definition stage into a mathematical model by using the information collected in the data collection phase. To do this we define two decision variables G and W to represent respectively the number of gizmos and widgets to be made and sold next month. Then the objective is to maximize total profits given by 10G+9W. There is a constraint corresponding to each of the three limited resources, which should ensure that the production of G gizmos and W widgets does not use up more of the corresponding resource than is available for use. Thus for resource 1, this would be translated into the following mathematical statement 0.7G+1.0W 630, where the lefthandside of the inequality represents the resource usage and the righthandside the resource availability. Additionally, we must also ensure that each G and W value considered is a nonnegative integer, since any other value is meaningless in terms of our definition of G and W. The completely mathematical model is:
Maximize {Profit = 10G+9W}, subject to

0.7G+1.0W 630

1.0G+(2/3)W 708

0.1G+0.25W 135

G, W 0 and integers.
This mathematical program tries to maximize the profit as a function of the production quantities (G and W), while ensuring that these quantities are such that the corresponding production is feasible with the resources available.
Model Solution: The fifth phase of the O.R. process is the solution of the problem represented by the model. This is the area on which a huge amount of research and development in O.R. has been focused, and there is a plethora of methods for analyzing a wide range of models. It is impossible to get into details of these various techniques in a single introductory chapter such as this; however, an overview of some of the more important methods can be found elsewhere in this handbook. Generally speaking, some formal training in operations research is necessary in order to appreciate how many of these methods work and the interested reader is urged to peruse an introductory text on O.R.; the section on "Further Reading" at the end of the chapter lists some good books. It is also worth mentioning that in recent years a number of software systems have emerged which (at least in theory) are "black boxes" for solving various models. However,
some formal education in O.R. methods is still required (or at least strongly recommended) before using such systems. From the perspective of the practitioner, the most important thing is to be able to recognize which of the many available techniques is appropriate for the model constructed. Usually, this is not a hard task for someone with some rudimentary training in operations research. The techniques themselves fall into several categories.
At the lowest level one might be able to use simple graphical techniques or even trial and error. However, despite the fact that the development of spreadsheets has made this much easier to do, it is usually an infeasible approach for most nontrivial problems. Most O.R. techniques are analytical in nature, and fall into one of four broad categories. First, there are simulation techniques, which obviously are used to anlyze simulation models. A significant part of these are the actual computer programs that run the model and the methods used to do so correctly. However, the more interesting and challenging part involves the techniques used to analyze the large volumes of output from the programs; typically, these encompass a number of statistical techniques. The interested reader should refer to a good book on simulation to see how these two parts fit together. The second category comprises techniques of mathematical analysis used to address a model that does not necessarily have a clear objective function or constraints but is nevertheless a mathematical representation of the system in question. Examples include common statistical techniques such as regression analysis, statistical inference and analysis of variance, as well as others such as queuing, Markov chains and decision analysis. The third category consists of optimumseeking techniques, which are typically used to solve the mathematical programs described in the previous section in order to find the optimum (i.e., best) values for the decision variables. Specific techniques include linear, nonlinear, dynamic, integer, goal and stochastic programming, as well as various networkbased methods. A detailed exposition of these is beyond the scope of this chapter, but there are a number of excellent texts in mathematical programming that describe many of these methods and the interested reader should refer to one of these. The final category of techniques is often referred to as heuristics. The distinguishing feature of a heuristic technique is that it is one that does not guarantee that the best solution will be found, but at the same time is not as complex as an optimumseeking technique. Although heuristics could be simple, commonsense, ruleofthumb type techniques, they are typically methods that exploit specific problem features to obtain good results. A relatively recent development in this area is socalled metaheuristics (such as genetic algorithms, tabu search, evolutionary programming and simulated annealing) which are general purpose methods that can be applied to a number of different problems. These methods in particular are increasing in popularity because of their relative simplicity and the fact that increases in computing power have greatly increased their effectiveness.
In applying a specific technique something that is important to keep in mind from a practitioner's perspective is that it is often sufficient to obtain a good solution even if it is not guaranteed to be the best solution. If neither resourceavailability nor time were an issue, one would of course look for the optimum solution. However, this is rarely the case in practice, and timeliness is of the essence in many instances. In this context, it is often more important to quickly obtain a solution that is satisfactory as opposed to expending a lot of effort to determine the optimum one, especially when the marginal gain from doing so is small. The economist Herbert Simon uses the term "satisficing" to describe this concept – one searches for the optimum but stops along the way when an acceptably good solution has been found.
At this point, some words about computational aspects are in order. When applied to a nontrivial, realworld problem almost all of the techniques discussed in this section require the use of a computer. Indeed, the single biggest impetus for the increased use of O.R. methods has been the rapid increase in computational power. Although there are still large scale problems whose solution requires the use of mainframe computers or powerful workstations, many big problems today are capable of being solved on desktop microcomputer systems. There are many computer packages (and their number is growing by the day) that have become popular because of their ease of use and that are typically available in various versions or sizes and interface seamlessly with other software systems; depending on their specific needs endusers can select an appropriate configuration. Many of the software vendors also offer training and consulting services to help users with getting the most out of the systems. Some specific techniques for which commercial software implementations are available today include optimization/ mathematical programming (including linear, nonlinear, integer, dynamic and goal programming), network flows, simulation, statistical analysis, queuing, forecasting, neural networks, decision analysis, and PERT/CPM. Also available today are commercial software systems that incorporate various O.R. techniques to address specific application areas including transportation and logistics, production planning, inventory control, scheduling, location analysis, forecasting, and supply chain management. Some examples of popular O.R. software systems include CPLEX, LINDO, OSL, MPL, SAS, and SIMAN, to name just a few. While it would clearly be impossible to describe herein the features of all available software, magazine such as OR/MS Today and IE Solutions regularly publish separate surveys of various categories of software systems and packages. These publications also provide pointers to different types of software available; as an example, the December 1997 issue of OR/MS Today (pages 6175) provides a complete resource directory for software and consultants. Updates to such directories are provided periodically. The main point here is that the ability to solve complex models/problems is far less of an issue today than it was a decade or two ago,
and there are plenty of readily available resources to address this issue.
We conclude this section by examining the solution to the model constructed earlier for our hypothetical production problem. Using linear programming to solve this model yields the optimal solution of G=540 and W=252, i.e., the production plan that maximizes profits for the given data calls for the production of 540 gizmos and 252 widgets. The reader may easily verify that this results in a profit of
$7668 and fully uses up all of the first two resources while leaving 18 units of the last resource unused. Note that this solution is certainly not obvious by just looking at the mathematical model – in fact, if one were "greedy" and tried to make as many gizmos as possible (since they yield higher profits per unit than the widgets), this would yield G=708 and W=0 (at which point all of the second resource is used up). However, the resulting profit of $7080 is about 8% less than the one obtained via the optimal plan. The reason of course, is that this plan does not make the most effective use of the available resources and fails to take into account the interaction between profits and resource utilization. While the actual difference is small for this hypothetical example, the benefits of using a good
O.R. technique can result in very significant improvements for large realworld problems.
Validation and Analysis: Once a solution has been obtained two things need to be done before one even considers developing a final policy or course of action for implementation. The first is to verify that the solution itself makes sense. Oftentimes, this is not the case and the most common reason is that the model used was not accurate or did not capture some major issue. The process of ensuring that the model is an accurate representation of the system is called validation and this is something that (whenever possible) should be done before actual solution. However, it is sometimes necessary to solve the model to discover inaccuracies in it. A typical error that might be discovered at this stage is that some important constraint was ignored in the model formulation – this will lead to a solution that is clearly recognized as being infeasible and the analyst must then go back and modify the model and resolve it. This cycle continues until one is sure that the results are sensible and come from a valid system representation.
The second part of this step in the O.R. process is referred to as post optimality analysis, or in layperson's terms, a "whatif" analysis. Recall that the model that forms the basis for the solution obtained is (a) a selective abstraction of the original system, and (b) constructed using data that in many cases is not 100% accurate. Since the validity of the solution obtained is bounded by the model's accuracy, a natural question that is of interest to an analyst is: "How robust is the solution with respect to deviations in the assumptions inherent in the model and in the values of the parameters used to construct it?" To illustrate this with our hypothetical production problem, examples of some questions that an analyst might wish to ask are, (a) "Will the optimum production plan change if the profits
associated with widgets were overestimated by 5%, and if so how?" or (b) "If some additional amount of Resource 2 could be purchased at a premium, would it be worth buying and if so, how much?" or (c) "If machine unreliability were to reduce the availability of Resource 3 by 8%, what effect would this have on the optimal policy?" Such questions are especially of interest to managers and decisionmakers who live in an uncertain world, and one of the most important aspects of a good O.R. project is the ability to provide not just a recommended course of action, but also details on its range of applicability and its sensitivity to model parameters.
Before ending this section it is worth emphasizing that similar to a traditional Industrial Engineering project, the end result of an O.R. project is not a definitive solution to a problem. Rather, it is an objective answer to the questions posed by the problem and one that puts the decisionmaker in the correct "ballpark." As such it is critical to temper the analytical solution obtained with common sense and subjective reasoning before finalizing a plan for implementation. From a practitioner's standpoint a sound, sensible and workable plan is far more desirable than incremental improvements in the quality of the solution obtained. This is the emphasis of this penultimate phase of the O.R. process.
Implementation and Monitoring: The last step in the O.R. process is to implement the final recommendation and establish control over it. Implementation entails the constitution of a team whose leadership will consist of some of the members on the original O.R. team. This team is typically responsible for the development of operating procedures or manuals and a timetable for putting the plan into effect. Once implementation is complete, responsibility for monitoring the system is usually turned over to an operating team. From an O.R. perspective, the primary responsibility of the latter is to recognize that the implemented results are valid only as long as the operating environment is unchanged and the assumptions made by the study remain valid. Thus when there are radical departures from the bases used to develop the plan, one must reconsider one's strategy. As a simple example with our production problem, if a sudden strike by the workforce causes a drastic reduction in the availability of labor (Resource 1, say), one must reconsider the plan completely to derive an alternative course of action. As a final word on implementation, it should be emphasized that a major responsibility of the operations research analyst is to convey the results of the project to management in an effective fashion. This is something that is unfortunately not emphasized sufficiently, and there are many instances of a successful study not being implemented because the details and the benefits are not conveyed effectively to management. While this is of course true of any project in general, it is especially significant with O.R. because of its mathematical content and its potential to not be fully understood by a manager without a strong quantitative background.


O.R. IN THE REAL WORLD
In this section some examples of successful real world applications of operations research are provided. These should give the reader an appreciation for the diverse kinds of problems that O.R. can address, as well as for the magnitude of the savings that are possible. Without any doubt, the best source for case studies and details of successful applications is the journal Interfaces, which is a publication of the Institute for Operations Research and the Management Sciences (INFORMS). This journal is oriented toward the practitioner and much of the exposition is in laypersons' terms; at some point, every practicing industrial engineer should refer to this journal to appreciate the contributions that O.R. can make. All of the applications that follow have been extracted from recent issues of Interfaces.
Before describing these applications, a few words are in order about the standing of operations research in the real world. An unfortunate reality is that O.R. has received more than its fair share of negative publicity. It has sometimes been looked upon as an esoteric science with little relevance to the realworld, and some critics have even referred to it as a collection of techniques in search of a problem to solve! Clearly, this criticism is untrue and there is plenty of documented evidence that when applied properly and with a problemdriven focus, O.R. can result in benefits that can be quite spectacular; the examples that follow in this section clearly attest to this fact.
On the other hand, there is also evidence to suggest that (unfortunately) the criticisms leveled against O.R. are not completely unfounded. This is because O.R. is often not applied as it should be – people have often taken the myopic view that O.R. is a specific method as opposed to a complete and systematic process. In particular, there has been an inordinate amount of emphasis on the modeling and solution steps, possibly because these clearly offer the most intellectual challenge. However, it is critical to maintain a problemdriven focus – the ultimate aim of an
O.R. study is to implement a solution to the problem being analyzed. Building complex models that are ultimately intractable, or developing highly efficient solution procedures to models that have little relevance to the real world may be fine as intellectual exercises, but run contrary to the practical nature of operations research! Unfortunately, this fact has sometimes been forgotten. Another valid criticism is the fact that many analysts are notoriously poor at communicating the results of an O.R. project in terms that can be understood and appreciated by practitioners who may not necessarily have a great deal of mathematical sophistication or formal training in O.R. The bottom line is that an O.R. project can be successful only if sufficient attention is paid to each of the seven steps of the process and the results are communicated to the endusers in an understandable form.
Some examples of successful O.R. projects are now presented.
Production Planning at Harris Corporation – Semiconductor Section: For our first application [1], we look at an area that is readily appreciated by every industrial engineer – production planning and due date quotation. The semiconductor section of Harris Corporation was for a number of years a fairly small business catering to a niche market in the aerospace and defense industries where the competition was minimal. However, in 1988 a strategic decision was made to acquire General Electric's semiconductor product lines and manufacturing facilities. This immediately increased the size of Harris Semiconductor's operations and product lines by roughly three times, and more importantly, catapulted Harris into commercial market areas such as automobiles and telecommunications where the competition was stiff. Given the new diversity of product lines and the tremendous increase in the complexity of production planning, Harris was having a hard time meeting delivery schedules and in staying competitive from a financial perspective; clearly, a better system was required.
In the orientation phase it was determined that the MRP type systems used by a number of its competitors would not be a satisfactory answer and a decision was made to develop a planning system that would meet Harris' unique needs – the final result was IMPReSS, an automated production planning and delivery quotation system for the entire production network. The system is an impressive combination of heuristics as well as optimizationbased techniques. It works by breaking up the overall problem into smaller, more manageable problems by using a heuristic decomposition approach. Mathematical models within the problem are solved using linear programming along with concepts from material requirements planning. The entire system interfaces with sophisticated databases allowing for forecasting, quotation and order entry, materials and dynamic information on capacities. Harris estimates that this system has increased ontime deliveries from 75% to 95% with no increase in inventories, helped it move from $75 million in losses to $40 million in profits annually, and allowed it to plan its capital investments more efficiently.
Gasoline Blending at Texaco: For another application to production planning, but this time in a continuous as opposed to discrete production environment, we look at a system in use at Texaco [2]. One of the major applications of O.R. is in the area of gasoline blending at petroleum refineries, and virtually all major oil companies use sophisticated optimization models in this area. At Texaco the system is called StarBlend and runs on networked microcomputers. As some background, the distillation of crude petroleum produces a number of different products at different distillation temperatures. Each of these may be further refined through cracking (where complex hydrocarbons are broken into simpler ones) and recombination. These various output streams are then blended together to form endproducts such as different grades of gasoline (leaded, unleaded, superunleaded etc.), jet fuel, diesel and heating oil. The planning problem is
very complex, since different grades of crude yield different concentrations of output streams and incur different costs, and since different endproducts fetch different revenues and use different amounts of refinery resources. Considering just one product – gasoline – there are various properties that constrain the blends produced. These include the octane number, lead and sulfur content, volatilities and Reid vapor pressure, to name a few. In addition, regulatory constraints impose certain restrictions as well.
As an initial response to this complex problem, in the early to mid 1980's Texaco developed a system called OMEGA. At the heart of this was a nonlinear optimization model which supported an interactive decision support system for optimally blending gasoline; this system alone was estimated to have saved Texaco about $30 million annually. StarBlend is an extension of OMEGA to a multi period planning environment where optimal decisions could be made over a longer planning horizon as opposed to a single period. In addition to blend quality constraints, the optimization model also incorporates inventory and material balance constraints for each period in the planning horizon. The optimizer uses an algebraic modeling language called GAMS and a nonlinear solver called MINOS, along with a relational database system for managing data. The whole system resides within a user friendly interface and in addition to immediate blend planning it can also be used to analyze various "whatif" scenarios for the future and for longterm planning.
FMS Scheduling at Caterpillar: For our third application we look at the use of a simulation model. This model was applied to derive schedules for a Flexible Manufacturing System (FMS) at Caterpillar, Inc. [3]. The interested reader may refer to any text on computer integrated manufacturing for details about FMSs; typically, they are systems of general purpose CNC machines linked together by an automated material handling system and completely controlled by computers. The FMS in question at Caterpillar had seven CNC milling machines, a fixturing station and a tool station, with material and tool handling being performed by four automated guided vehicles (AGVs) traveling along a oneway guided wire path. FMSs can provide tremendous increases in capacity and productivity because of the high levels of automation inherent in them and their potential to manufacture a wide variety of parts. On the other hand, this comes with a price; these systems are also very complex and the process of planning and scheduling production on an FMS and then controlling its operation can be a very difficult one. The efficiency of the scheduling procedure used can have a profound effect on the magnitude of the benefits realized.
At Caterpillar, a preliminary analysis showed that the FMS was being underutilized and the objective of the project was to define a good production schedule that would improve utilization and free up more time to produce additional parts. In the orientation phase it was determined that the environment was much too complex to represent it
accurately through a mathematical model, and therefore simulation was selected as an alternative modeling approach. It was also determined that minimizing the makespan (which is the time required to produce all daily requirements) would be the best objective since this would also maximize as well as balance machine utilization. A detailed simulation model was then constructed using a specialized language called SLAM. In addition to the process plans required to specify the actual machining of the various part types, this model also accounted for a number of factors such as material handling, tool handling and fixturing. Several alternatives were then simulated to observe how the system would perform and it was determined that a fairly simple set of heuristic scheduling rules could yield near optimal schedules for which the machine utilizations were almost 85%. However, what was more interesting was that this study also showed that the stability of the schedule was strongly dependent on the efficiency with which the cutting tools used by the machines could be managed. In fact, as tool quality starts to deteriorate the system starts to get more and more unstable and the schedule starts to fall behind due dates. In order to avoid this problem, the company had to suspend production over the weekends and replace wornout tools or occasionally use overtime to get back on schedule. The key point to note from this application is that a simulation model could be used to analyze a highly complex system for a number of whatif scenarios and to gain a better understanding of the dynamics of the system.

SUMMARY
This chapter provides an overview of operations research, its origins, its approach to solving problems, and some examples of successful applications. From the standpoint of an industrial engineer, O.R. is a tool that can do a great deal to improve productivity. It should be emphasized that
O.R. is neither esoteric nor impractical, and the interested
I.E. is urged to study this topic further for its techniques as well as its applications; the potential rewards can be enormous.

REFERENCES


V.K. Kapoor: Operations Research, Sultan Chand & Sons, 2008.

Hamdy.A.Taha: Operations Research, Eighth Edition, Prince Hall of India, New Delhi, 2008.

Kanti Swarup, Gupta & P.K.Man Mohan: Operations Research, Sultan Chand & Sons, 2007

Panneerselvam: Operations Research, 2004.

H.S.Kasana, K.D.Kumar: Introductory Operations Research, Springer, 2003.

Ravindran, Phillips, Solberg: Operations Research, Second Edition, John Wiley & Sons, 2007.

Wayne L.Winston: Operations Research, Fourth Edition, Thomson Brooks/Cole, 2003.

P.Mariappan: Operations Research, New Century Book House (P) Limited, 2003.