INTRODUCTION TO ENTERPRISE RESOURCE PLANNING SYSTEMS IN MANUFACTURING
INTRODUCTION
Enterprise resource planning (ERP) is a class of commercially developed software applications that integrate a vast array of activities and information to support tactical-level operations and operations planning for an industrial enterprise. The term ERP refers to the software and not to the related business processes. However, as software, it enables better execution of certain processes. Although often presented as a single package, an ERP system is an envelope around numerous applications and related information. For manufacturers, those applications typically support the operations pro- cesses of materials sourcing, manufacturing planning, and product distribution. To its end users, an individual application of an ERP system may appear seamless; however, to those who procure, implement, and / or maintain ERP systems, they are complex software systems that require varying levels of customization and support both centrally and across applications. While ERP systems are commercial applications developed by individual vendors, they can hardly be considered off-the- shelf. They are part of a continuing trend of outsourcing IT solutions in which part of the solution is bought, part is configured, and part is built from scratch. In general, as the scope and complexity of integrated applications have increased from systems supporting a single business unit to systems supporting an entire enterprise and its relationships with business partners, the portions of an IT solution that are bought and configured have increased while the percentage of custom-built software has decreased. Given their broad organizational and functional scope, ERP systems are unlike any other contemporary commercial manufacturing applications. They provide ‘‘transaction manage- ment,’’ both from the business perspective and from a database perspective. Additionally, they provide a basic level of decision support. Optionally, they enable development of software for higher levels of decision support, which may be offered by ERP vendors or third-party vendors. It is clear that ERP, as a subject, is very complex. Its use marries technology, business practices, and organizational structures. The purpose of this chapter is to present a high-level view of ERP in order to frame a discussion of technological challenges and research opportunities for improving ERP interoperability. Although ERP is relevant to many types of industries (e.g., goods and services) and organizations (e.g., for-profit and not-for-profit), the discussion in this chapter is limited to ERP in manufacturing enterprises. More specifically, the focus of this chapter is ERP that supports the principal operations of a manufacturing enterprise: planning, procuring, making, and delivering products. An ERP system may support other enterprise functions, such as finance management, human resource management, and possibly sales and marketing activities. Detailed analysis of those functions is beyond the scope of this chapter; however, the linkages of those functions with manufacturing-specific functions are not.
This overview looks at ERP by itself and as part of a larger entity (Figure 1). Section 2 discusses ERP internals such as core functions, implementation elements, and technology issues. Additionally,
Section 2 identifies critical integration points for ERP and other applications within manufacturing enterprises. Section 3 discusses ERP and its relationship to three larger entities, namely the U.S. economy, supply chains, and individual manufacturers. Section 4 presents issues and possible reso- lutions for improving ERP performance and interoperability.
This chapter is the result of a two-year study funded by two programs at the National Institute of Standards and Technology: the Advanced Technology Program’s Office of Information Technology and Applications and the Manufacturing Systems Integration Division’s Systems for Integrating Man- ufacturing Applications (SIMA) Program.
The concepts presented in this chapter were gathered from a variety of sources, including literature reviews, manufacturing industry contacts, ERP vendor contacts, consultants specializing in the ap- plications of IT to manufacturing, relevant professional and trade associations, and standards orga- nizations.
Major Business Functions in Manufacturing Enterprises
Manufacturers typically differentiate themselves from competitors along the three major business functions through which they add value for their customers. Customer relationship management (CRM), the first dimension of competitive advantage, seeks to add value for customers through those processes that involve direct contact with customers before, during, and after sales. The idea is to understand the prevalent needs and concerns of individual customers and groups of customers. Prod- uct development, the second dimension of competitive advantage, focuses on product—what and how to produce an object to satisfy the customer’s want. Operations, the third dimension of competitive advantage, focuses on satisfying demand—how much to make, when to make, and where to make— by producing and delivering products in an effective and efficient manner.
Needs, wants, and demands are basic concepts underlying modern, market-based economies (Kot- ler and Armstrong 1999). Needs beget wants, which beget demand. Needs are states of felt depri- vation. They are a basic part of our human condition and are physical, social, and individual in nature. The customer’s needs include product capabilities, product servicing, user instruction, and business relationships. Wants are the forms taken by human needs as shaped by culture and individual personality. They are described in terms of objects that will satisfy needs. Demands are human wants that are backed by buying power. Identifying needs and translating them into wants in terms of product and process definitions are the objectives of product development. Satisfying demand, given supply conditions as well as product and process definitions, is the objective of operations. This high- level partitioning of manufacturing business functions into CRM, product development, and opera- tions has growing acceptance in manufacturing and related industries (Hagel and Singer 1999). This acceptance has been fostered by the realization that the underlying activities of these high-level functions are those that add value for the customer.
The complex activities of product development seek to satisfy customer want by translating the abstract to the physical through the product development process. As such, in commercial manufac- turing enterprises, product development typically starts with an analysis of market opportunity and strategic fit and, assuming successful reviews through intermediate phases, ends with product release. Among other things, product release serves as a signal to operations to begin production and distri- bution as the necessary design and manufacturing engineering specifications are ready for execution in a production environment.
Operations, on the other hand, consists of processes for satisfying customer demand by trans- forming products—in raw, intermediate, or final state—in terms of form, location, and time. To accomplish this objective both effectively and efficiently—and thus meet specific, customer- focused, operational objectives—a manufacturing enterprise must have timely and accurate information about expected and real demand as well as expected and real supply. A manufacturer then considers this information on supply and demand with the current and expected states of its enterprise. It is ERP that allows a manufacturer to monitor the state of its enterprise—particularly the current and near- term expected states. In fact, ERP systems often serve as the cornerstone in the emerging information architectures that support balancing external and internal supply and demand forces. ERP systems play both direct and indirect roles in this trend among manufacturing enterprises towards a synchro- nized, multilevel, multifacility supply chain planning hierarchy.
Manufacturing Operations Planning
Figure 2 illustrates the emerging synchronized, multilevel, multifacility supply chain planning hier- archy with ERP as its foundation. The goal of this architecture is to enable more efficient and effective execution across plants, distribution systems, and transportation systems. Independently, these plan- ning activities focus on the strategic, the tactical, and the operational (i.e., execution) levels. Collec- tively, they support the translation of strategic objectives into actions on the plant floor, in warehouses, and at shipping points throughout the extended enterprise. In addition, they provide top management with up-to-date, synchronized information regarding the state of the entire enterprise.
This synchronization is accomplished by transforming information in a meaningful way from one level within the supply chain planning hierarchy to the next. At the strategic level, top management evaluates numerous factors to determine the design or redesign of the supply chain network as well as time-independent sourcing, production, deployment, and distribution plans. These factors typically include the enterprise’s business philosophy as well as company, market, technological, economic, social, and political conditions. Supply chain planning at the strategic level involves ‘‘what if’’ anal- ysis particularly with respect to the first three factors: business philosophy, company conditions, and market conditions. A business philosophy might specify maximizing net revenues or return on assets. Assessment of company conditions considers existing and potential aggregates of fixed (i.e., plant), financial, and human resources. When evaluating market conditions, top management analyzes ag- gregate product / part demand as well as the anticipated capacity of suppliers and transportation chan- nels—also in aggregate terms. Optimization at this level, which usually employs mathematical programming methods, typically yields the location, size, and number of plants, distribution centers, and suppliers as well as product and supply volumes.
Supply chain operations planning at the tactical level determines the flow of goods over a specific time horizon. Mathematical programming methods yield time-dependent integrated sourcing, pro-
duction, deployment, and distribution plans typically designed to satisfy a financial management objective such as minimizing total supply chain costs or maximizing net revenues by varying product mix. Often, once these plans are established, a more detailed level of tactical planning occurs to optimize supply, production, and distribution independently. Frequently, the objective at this lower level of tactical planning is to minimize costs related to inventories and / or major equipment change- overs.
Supply chain planning at the operational level is, in essence, supply scheduling that occurs on a facility-by-facility basis. As such, separate but coordinated schedules are generated for plants, ware- houses, distribution centers, and vehicle systems. Planning at this level differs from tactical and strategic levels in that demand actually exists—that is, orders have been placed. These orders need to be scheduled based on the immediate state of resources (i.e., materials, equipment, and labor). The diverse nature of facilities means that the specifics of optimization vary widely at this level, but the objective typically is to maximize throughput in a given facility.
Individually, these layers serve to separate concerns and enable the definition of tractable planning problems for which mathematical and managerial solutions can be obtained. Collectively, these layers of supply chain planning enable manufacturing enterprises more effectively and efficiently to balance supply, resources, and demand. The upper layers buffer the lower layers from sudden shifts in the market, thus allowing for smoother changes in the enterprise’s plants, distribution channels, and transportation systems.
Partitioning the Domain of Manufacturing
The domain of manufacturing is in fact an aggregate of many subdomains of many types. There is no single correct method for decomposing that complex and dynamic aggregate. The method of decomposition depends on the particular objective at hand. Generally speaking, there are four com- mon approaches to partitioning the manufacturing domain. Each looks at a different aspect of the manufacturing enterprise:
1. Nature of the product: This approach categorizes manufacturing industries by the general nature of the product itself—fertilizers, pharmaceuticals, metals, automotive parts, aircraft, etc. This is the approach used by industry classification systems such as the North American Industry Classification System (NAIC) (Office of Management and Budget 1997) and its pre- decessor, the Standard Industrial Classification (SIC) (Office of Management and Budget 1988). In general, this approach is a good mechanism for characterizing market communities, and thus economic estimators, but it is not a particularly good mechanism for characterizing ERP requirements or planning approaches.
2. Nature of the customer: Because most manufacturing output is consumed by other industries, many manufacturers are part of the supply chains ending in original equipment manufacturers (OEMs) in a single major industry: automotive, aerospace, shipbuilding, household appliances, computers, etc. The members of the chain produce different kinds of products, using different processes, with different business behaviors, but the behavior of the supply chain itself often is dominated by the demands of the OEMs.
3. Nature of the process: This approach characterizes a domain by the organization of the man- ufacturing facility and the general nature of the manufacturing processes it employs: continuous process, assembly line, discrete batch, job shop, and construction. There is some correlation between the process type and the product type, in that most manufacturers of a given product type tend to use a particular process organization. In general, process type strongly influences the manufacturing-specific aspects of ERP, including both information capture and planning approaches. On the other hand, large manufacturers often use several different process styles for different products and different components of larger products.
4. Nature of the business in terms of customer orders: This characterization includes make-to- stock, make-to-order, assemble-to-order, and engineer-to-order. It has a great deal to do with what the detailed business operations are and how operational and tactical planning is done. Clearly this categorization has a tremendous influence on the ERP requirements and on the behavior of the enterprise in its supply chain. More than any other, this characterization de- termines the nature of the delivery activities and the dependence on supplier relationships.
These last two categories, as differentiators for ERP, warrant more detailed discussion.
Nature of Process
Continuous process refers to a facility in which products are made by an essentially continuous flow of material through some set of mixing, state transformation, and shaping processes into one or more final products. The final form may itself be intrinsically discrete, or it may be discretized only for packaging and shipment. Examples are wet and dry chemicals, foods, pharmaceuticals, paper, fibers, metals (e.g., plate, bar, tubing, wire, sheet), and pseudocontinuous processes such as weaving, casting, injection molding, screw machines, and high-volume stamping.
Assembly line refers to a facility in which products are made from component parts by a process in which discrete units of product move along an essentially continuous line through a sequence of installation, joining, and finishing processes. Examples are automobiles, industrial equipment, small and large appliances, computers, consumer electronics, toys, and some furniture and clothing.
Discrete batch, also called intermittent, refers to a facility in which processes are organized into separate work centers and products are moved in lots through a sequence of work centers in which each work center is set up for a specific set of operations on that product and the setup and sequence is specific to a product family. This describes a facility that can make a large but relatively fixed set of products but only a few types of product at one time, so the same product is made at intervals. This also describes a facility in which the technology is common—the set of processes and the ordering is relatively fixed, but the details of the process in each work center may vary considerably from product to product in the mix. Examples include semiconductors and circuit boards, composite parts, firearms, and machined metal parts made in quantity.
Job shop refers to a facility in which processes are organized into separate work centers and products are moved in order lots through a sequence of work centers in which each work center performs some set of operations. The sequence of work centers and the details of the operations are specific to the product. In general, the work centers have general-purpose setups that can perform some class of operations on a large variety of similar products, and the set of centers used, the sequence, the operations details, and the timing vary considerably over the product mix. Examples include metal shops, wood shops, and other piece-part contract manufacturers supporting the auto- motive, aircraft, shipbuilding, industrial equipment, and ordnance industries.
Construction refers to a manufacturing facility in which the end product instances rarely move; equipment is moved into the product area and processes and component installations are performed on the product in place. The principal examples are shipbuilding and spacecraft, but aircraft manu- facture is a hybrid of construction and assembly line approaches.
Nature of the Business in Terms of Customer Orders
Make-to-stock describes an approach in which production is planned and executed on the basis of expected market rather than specific customer orders. Because there is no explicit customer order at the time of manufacture, this approach is often referred to as a push system. In most cases, product reaches retail outlets or end customers through distribution centers and manufacturing volumes are driven by a strategy for maintaining target stock levels in the distribution centers.
Make-to-order has two interpretations. Technically, anything that is not made-to-stock is made- to-order. In all cases there is an explicit customer order, and thus all make-to-order systems are described as pull systems. However, it is important to distinguish make-to-demand systems, in which products are made in batches, from option-to-order systems, in which order-specific features are installed on a product-by-product basis. The distinction between make-to-demand batch planning and on-the-fly option selection using single setup and prepositioning is very important to the ERP system.
A make-to-demand manufacturer makes fixed products with fixed processes but sets up and ini- tiates those processes only when there are sufficient orders (i.e., known demand) in the system. This scenario may occur when there is a large catalog of fixed products with variable demand or when the catalog offers a few products with several options. The distinguishing factor is that orders are batched and the facility is set up for a run of a specific product or option suite. The planning problem for make-to-demand involves complex trade-offs among customer satisfaction, product volumes, ma- terials inventories, and facility setup times.
Option-to-order, also called assemble-to-order, describes an approach in which production is planned and executed on the basis of actual (and sometimes expected) customer orders, in which the product has some predefined optional characteristics which the customer selects on the order. The important aspects of this approach are that the process of making the product with options is pre- defined for all allowed option combinations and that the manufacturing facility is set up so the operator can perform the option installation on a per-product basis during manufacture. This category also applies to a business whose catalog contains a family of fixed products but whose manufacturing facility can make any member of the family as a variant (i.e., option) of a single base product. The option-to-order approach effects the configuration of production lines in very complex ways. The simplest configurations involve prepositioning, in which option combinations occur on the fly. More complex configurations involve combinations of batching and prepositioning.
Engineer-to-order describes an approach in which the details of the manufacturing process for the product, and often the product itself, must be defined specifically for a particular customer order and only after receipt of that order. It is important for other business reasons to distinguish contract engineering, in which the customer defines the requirements but the manufacturer defines both the product and the process, from contract manufacturing, in which the customer defines the product details and the manufacturer defines the process details. But the distinction between contract engi- neering and contract manufacturing is not particularly important for ERP, as long as it is understood that both are engineer-to-order approaches. In these scenarios, some set of engineering activities must take place after receipt of customer order and before manufacturing execution, and then certain aspects of manufacturing planning can begin.
Comments
Post a Comment