Group Technology and Productivity

Introduction

Group technology (GT) is gaining increasing popularity from manufacturers because of its many applications for boosting productivity. GT is an approach to manufacturing that seeks to maximize production efficiencies by grouping similar and recurring problems or tasks. An important part of GT is the use of a code that—like a library reference system—serves as an index to characteristics in manufacturing, engineering, purchasing, resource planning, and sales to improve productivity in each of these areas.

Cost of introducing a new part includes expenses for design, planning and control, and tools and fixtures. Clearly, if a company can reduce the number of new parts it needs, it would realize large cost savings. But this saving comes with an additional task: the identification of parts that can be used with or without modification to meet the designer’s need. A “group technology” manufacturing data base offers great assistance in this identification process.

Group technology (GT) is a concept that currently is attracting a lot of attention from the manufacturing community. The essence of GT is to capitalize on similarities in recurring tasks in three ways:

• By performing similar activities together, thereby avoiding wasteful time in changing from one unrelated activity to the next.

• By standardizing closely related activities, thereby focusing only on distinct differences and avoiding unnecessary duplication of effort.

• By efficiently storing and retrieving information related to recurring problems, thereby reducing the search time for the information and eliminating the need to solve the problem again.

The management of manufacturing technologies represents a vital component in the competitiveness of the industry in general, one that should play a more important role in the formulation of strategic plans. For this to happen, general management must become more familiar with emerging and promising technologies. In this article we discuss several such technologies, all tied together by group technology, and identify their wide-reaching applications to all areas of business operations. We describe the potential benefits that can be achieved as well as common implementation problems.

The Meaning of Group Technology

Group Technology states that managers should exploit similarities and achieve efficiencies by grouping similar problems. In most cases, a prerequisite for the recognition of similarities is a system by which the objects of interest can be classified and coded (that is, assigned symbols representing relevant information). As an analogy, books in a library catalog are classified and coded in such a way that one can easily find all books written by a particular author, covering a certain topic, or sharing the same title.

In design engineering, parts can be classified by geometric similarities using codes which contain design attributes. The purpose could be to retrieve all parts with certain features, such as rotational parts with a length-to-diameter ratio of less than 2. If one of these fits the need at hand, the engineers can thereby avoid having to design a new part.

Similarities between parts, captured in the GT code, can in like manner be used by manufacturing engineering, manufacturing, purchasing, and sales. For example, a manufacturer can drastically reduce the time and effort spent deciding how a part should be produced if this information is available for a similar part.

A GT data base is a computerized filing system that speeds up the retrieval of parts information, facilitates the design process, improves the accuracy of process planning, aids in the creation and operation of manufacturing cells, and enhances the communication between functional areas.

An early use of GT was documented in the Soviet Union in the 1940s. It has since been implemented in many European and Asian countries, mainly in the manufacturing area. Interest among U.S. manufacturers took root in the mid-1970s, and by now many large corporations (John Deere, Caterpillar, Lockheed, General Electric, Black & Decker, and Cincinnati Milacron are a few) have taken advantage of GT or are planning GT programs.

The expansion of computer capabilities and the availability of software obviously have abetted the growth of GT applications. Storing and retrieving codes with 20 or 30 characters are unthinkable without the aid of a computer. But with advanced employment of the computer in any area of production operations also comes the need for coding and classification as a way to integrate tasks and even organizational units. This is the reason why many experts see GT as the missing link between CAD and CAM (computer-aided design and computer-aided manufacturing) and thereby as an important building block for CIM (computer-integrated manufacturing).

Classification & Coding

GT is predominantly applied to purchased items and fabricated parts. We will concentrate our discussion on these groups.

Group Technology, Source: https://www.industryemea.com/

When engineers are classifying parts and assigning those with closely related attributes to a particular family, they can determine similarities between items in several ways. From a design standpoint, for example, similarity can mean closely related geometric shapes and dimensions. From a manufacturing point of view, similarity between two parts means that they are processed through the factory in the same or almost the same way. Of course, parts that look alike are not always produced in the same way (it depends on variations in raw materials, tolerances, dimensions, and so on), while parts that are routed through the same machines can be quite dissimilar in geometric form.

Numerous coding systems have been developed all over the world by university researchers and consulting firms and also by corporations for their own use. A handful of commercial systems is available on the U.S. market. Of all these, many have a short range of applications, such as coding sheet metals or forgings only. Modern systems, however, are often computerized, and coding takes place by having the planner work in a conversational mode with the computer, responding to a series of questions on the CRT screen.

Once the parts have been coded, classification, which simply means the grouping of parts with similar characteristics, follows. For example, one family could consist of all parts with a 2 in the third position of the code. Another could be of those with a 3 in the third position and a 2 in the fourth. It is easy to see that with a code length of 30 digits, it is possible to create a very large number of families.

Applying GT

Although many areas of business operations can benefit from GT, manufacturing, the original application area, continues to be the place where GT is most widely practiced. Two important tasks in manufacturing planning and manufacturing engineering are scheduling and process planning. Job scheduling sets the order in which parts should be processed and can determine expected completion times for operations and orders. Process planning, on the other hand, decides the sequence of machines to which a part should be routed when it is manufactured and the operations that should be performed at each machine. Process planning also encompasses tool, jig, and fixture selection as well as documentation of the time standards (run and setup time) associated with each operation. Process planning can directly affect scheduling efficiency and, thus, many of the performance measures normally associated with manufacturing planning and control.

1. in production planning & control

Grouping parts with similar manufacturing characteristics into families will reduce the time spent on setups of parts and tools. In small-to medium-batch manufacturing, striving for setup reduction is most important. This type of parts production usually is carried on in a job shop environment where general purpose machines are grouped according to function, such as lathes in one cluster and grinders in another. Job shops also usually have high work-in-process inventories, long lead times, and an extremely low productive use of the time a part spends on the shop floor (normally no more than 5% of the total shop time). The following are among the ways GT can be carried out in production planning.

2. Sequencing of parts families

The simplest—and a highly informal—application of GT in a job shop setting is to sequence similar parts on a machine. This procedure, followed daily by foremen in most machine shops, often means overriding formal dispatch lists, which are made up with no consideration of efficiency. The saved setup time from running two or more related parts in a row can be converted to productive time. A more sophisticated application is the creation of families of parts (using the GT code) and the dedication of machines for exclusive processing of families. This approach has several advantages. First, there are fewer interfering flows of material at each machine. Second, setup time is reduced since common tooling and fixtures can be developed to handle all members of each family processed at the work station. Third, the quality of parts can be improved, since the variety of parts flowing through the work station has been reduced.

3. Cellular production

The most advanced GT application is through the creation of manufacturing cells. A cell is a collection of machine tools and materials-handling equipment grouped to process one or several part families. Preferably, parts are completed within one cell. (The Japanese make much use of such cells, but apparently without formal classification and coding systems.) The advantages of cellular manufacturing are many, especially when the cells are designed with one dominant materials flow and with a fixed conveyor system connecting the work stations. A cell represents a hybrid production system, a mixture of a job shop producing a large variety of parts and a flow shop dedicated to mass production of one product. Exhibit II illustrates the difference between a job shop, based on a functional layout, and a cell shop.

The allocation of equipment to a subset of parts will reduce interference, improve quality, make materials handling more efficient, cut setup and run times, and therefore trim inventories and shorten lead times. Shortening parts manufacturing lead times can reduce the response time to customer orders and thus lead to smaller finished-goods inventories as well. These benefits are likely to be greater with a physical rearrangement of machinery into cells.

Cellular manufacturing offers other advantages too. The factory layout change has organizational and behavioral implications. Otis Engineering, for example, achieved a more efficient use of supervisory personnel, and equipment operators gained flexibility and thereby job enrichment. Otis also established centralized tool and gauge storage for the cell to permit easier access to the tools and better tool scheduling.

Managers can simplify production planning and control by considering the cell as one planning point for which capacity planning can be performed and to which jobs can be released. Cells commonly have more machines than operators, which means that the operators must balance the load in the cell. This, of course, represents a decentralization of tasks, requiring operators to handle several machine tools and processes. Further, with cellular manufacturing, tracing a part to its origin is easy, which facilitates accountability for quality. Management can exploit this by assigning the responsibility for quality inspection to the cell operators.

GT code generation over internet, Source: https://isr.umd.edu/

Manufacturing cells also change the tasks production planners, schedulers, and manufacturing engineers perform and, most dramatically, alter the role of the foreman. Where before he or she was responsible for only one process, the foreman now supervises production of an entire part. This, too, affects accountability. In a job shop environment it is always possible to pass the buck by blaming other foremen for not having parts ready on time. With cells, timely completion becomes a responsibility solely of the cell foreman.

By mechanizing and automating the materials handling and the manufacturing process, engineers can create unmanned cells based on GT principles. A robotic work cell designed to process a small set of part families, for example, consists of computer-controlled machine tools located around one or more materials-handling robots. Since there is no fixed machine sequence, this type of cell can be quite flexible. A somewhat less flexible and often much larger cell, designed for higher volumes and more specialized parts, is called a flexible manufacturing system.

4. in process planning

Some of the largest productivity gains have been reported in the creation of process plans that determine how a part should be produced. With computer-aided process planning (CAPP) and GT it is possible to standardize such plans, reduce the number of new ones, and store, retrieve, edit, and print them out very efficiently.

Process planning normally is not a formal procedure. Each time a new part is designed, a process planner will look at the drawing and decide which machine tools should process the parts, which operations should be performed, and in what sequence.

There are two reasons why companies often generate excess process plans. First, most companies have several planners, and each may come up with a different process plan for the very same part. Second, process planning is developed with the existing configuration of machine tools in mind. Over time, the addition of new equipment will change the suitability of existing plans. Rarely are alterations to old process plans made. One company reportedly had 477 process plans developed for 523 different gears. A close look revealed that more than 400 of the plans could be eliminated. Another company used 51 machine tools and 87 different process plans to produce 150 parts. An investigation determined that these parts could be produced on only 8 machines via 31 process plans. Process planning using CAPP can avoid these problems.

Process planning with CAPP takes two different forms:

With variant-based planning, one standardized plan (and possibly one or more alternate plans) is created and stored for each part family. When the planner enters the GT code for a part, the computer will retrieve the best process plan. If none exists, the computer will search for routings and operations sequences for similar parts. The planner can edit the scheme on the CRT screen before printout.

With generative planning, which can but does not necessarily rely on coded and classified parts, the computer forms the process plan through a series of questions the computer poses on the screen. The end product is also a standardized process plan, which is the best plan for a particular part.

The variant-based approach relies on established plans entered into the computer memory, while the generative technique creates the process plans interactively, relying on the same logic and knowledge that a planner has. Generative process planning is much more complex than variant-based planning; in fact, it approaches the art of artificial intelligence. It is also much more flexible: by simply changing the planning logic, for instance, engineers can consider the acquisition of a new machine tool. With the variant-based method, the engineers must look over and possibly correct all plans that the new tool might affect.

CAPP permits creation and documentation of process plans in a fraction of the time it would take a planner to do the work manually and vastly reduces the number of errors and the number of new plans that must be stored. When you consider that plans normally are handwritten and that process planners spend as much as 30% of their time preparing them, CAPP’s contribution of standardized formats for plans and more readable documents is important. CAPP, in effect, functions as an advanced text editor. Furthermore, it can be linked with an automated standard data system that will calculate and record the run times and the setup times for each operation.

CAPP can lead to lower unit costs through production of parts in an optimal way. That is, cost savings come not only via more efficient process planning but also through reduced labor, material, tooling, and inventory costs. 

GT can help in the creation of programs that operate numerically controlled (NC) machinery, an area related to process planning. For example, after the engineers at Otis Engineering had formed part families and cells, the time to produce a new NC tape dropped from between 4 and 8 hours to 30 minutes. The company thereby improved the potential for use of NC equipment on batches with small manufacturing quantities.⁷

5. in parts design

GT coding of parts is useful for the efficient retrieval of previous designs as well as for design standardization. These features help speed up the design process and curb design proliferation.

It is not unusual for a company to find several versions of basically the same part during a preliminary investigation of the part population. The parts can serve the same function but differ in terms of tolerances, radii, and so on. 

Design proliferation of this kind occurs because of difficulties with design retrieval. While a part similar to the one that is needed may already exist, the designer has neither a system nor the patience to find it. It is easier to create a new part, which then means that a new part number must be assigned, a new process plan made up, new tools designed, and so on.

The aim of design standardization is to reduce variations, to make the parts efficiently, and to require justification for deviations from norms. Standardization does not mean that all parts with the same function must be identical. It does mean, however, that norms are established for tolerances, dimensions, angles, and other specifications. Setting these norms should be done with both manufacturing and design considerations in mind, bridging the gap between these two areas and making design engineers more aware of manufacturing costs and restrictions.

A GT coded parts population simplifies the cumbersome job of sifting through old drawings to find an already designed part. The designer can enter on a CRT a partial code describing the main characteristics of the needed part. The computer will then search the GT data base for all items with the same code and list them on the screen. The designer can go through the specifications of each part and select one that fits or can be modified. With modern computer graphics, each designed part can be displayed on the CRT so that the designer can inspect it. Once a part design has been selected or “edited,” the designer can make the actual drawing manually or by computer.

6. in other areas

GT can also be applied in purchasing. Relying on the GT coding of purchased components and raw materials and on information from the production planning system, a purchasing manager can obtain statistics not directly available with a traditional parts numbering system. GT can help reduce proliferation of purchases of different kinds of parts, for example, by identifying components that serve the same function. It can also list identical parts for which designers have specified different brands. Companies that reduce the number of different parts they order and the suppliers they do business with can use the increased volume as leverage to negotiate better deals.

Another interesting application is in sales. The same company received a request for immediate delivery of an engine bolt that was not a stock item. A search of the GT data base, however, turned up a substitute part that fit the customer’s need and could be delivered right away.

GT can also be used for cost estimation. A company that needs product cost estimates for bidding purposes, for example, can tentatively code the required parts and then search the GT data base. For parts falling into established families, standard cost data might already exist. If not, the CAPP system can help to determine the processes needed to manufacture the part, thereby arriving at cost data. Several companies have found that GT-generated cost estimates can be constructed more quickly and with greater accuracy than those made by traditional methods. The approach is also helpful during the design process to help select components that will lower the total cost of the proposed product.

GT can also assist in determining the economic consequences of anticipated changes in materials cost. Assume, for example, that the price of an already expensive alloy is expected to rise. With GT coding, a list of all parts that use this alloy can be produced within minutes, permitting a swift assessment of how the increased purchasing cost will affect the manufacturing cost of products made with the parts.

Implementing GT

GT is a philosophy calling for simplicity and standardization. Any serious attempt to take full advantage of it begins with the selection of one or more coding systems (each type of material can conceivably have its own coding system) and the subsequent coding of the material. As with any formal information system that requires changing ingrained methods and old procedures, GT cannot be casually decided on or instantaneously implemented. A GT program could require two or three years to install and will have far-reaching ramifications inside the organization, particularly if cellular manufacturing is instituted.

Resistance to change, of course, is a universal problem in any organization. Resistance can take different forms, depending on the employee’s perception of job status and security, understanding of the new situation, and ability and willingness to adapt. In the case of GT, some examples illustrate the range of potential problems.

• The mandate for designers to reduce the number of new parts can conflict with a company’s long-standing evaluation and reward system. In one company, designers had been evaluated on the number of new drawings they created. Instituting the variety-reduction concept, therefore, necessarily meant changing the incentive system. (Installing CAD by itself can lead to a surge in new parts, simply because of the speed with which new designs can be produced.)

• In manufacturing, problems commonly stem from the changing roles of operators and the new areas of responsibility for supervisors. Working as a team and participating in decision making puts employees in a new sociological setting. Operators should be able to move from work station to work station and to perform quality inspection. This requires additional skills and constitutes a new job design. Both workers and labor union representatives often resist these changes. The foreman’s role also expands to cover many functions. The foreman must, therefore, know several manufacturing processes instead of only one and be responsible for the completion of the whole part and not for only a single operation.

Extensive education about GT concepts, hands-on training, and the early involvement of the affected individuals are the best ways to implement new work roles. Selling the idea of GT cells to labor unions can require great efforts, including restructuring payment systems. Personnel policies and training systems must change as well. Because of the different requirements of working in a cell, companies commonly rely on volunteers when forming teams. The Alfa-Laval company in Sweden lets the workers form their own teams and then trains them by simulating the movements and coordination inside the manufacturing cell on a scale model.

• Production planners and schedulers are also directly affected by cellular manufacturing. Once a cell has been established, parts not belonging to the appropriate family must be routed elsewhere. Some companies have found that planners have a tendency to break the family rule: they schedule a part to the cell containing machines they consider to be the most efficient for its production or to cells that have become so efficient that they seem underused. It is difficult to keep the new system up, and lapses like these can destroy the system’s integrity and lead to backsliding to old ways. Regular monitoring should accompany the period immediately following a changeover, when these lapses are most likely to occur.

• The most common problem companies face in the area of coding and classification stems from the inability of codes to describe the material adequately. Some companies, for example, found that their coding systems were suitable for design but not for manufacturing purposes. A manager of one company commented that its coding system could not handle electrical parts. In fact, about two-thirds of the companies with externally developed systems had made modifications to suit their own needs. This finding reflects two facts. First, codes to handle both design and manufacturing are relatively new, and second, companies often need to develop procedures that reflect their own ways of doing things.

• One direct implication of cellular manufacturing is that the more rigid, flow-oriented system with drastically reduced work in process requires a stronger emphasis on machine maintenance. Cell formation, however, also creates a visibility that does not exist in job shops. This visibility makes it easier for managers and supervisors to identify load-balancing problems in the cells. When a manufacturing cell is designed, one obvious goal is to achieve a high utilization of all machines in the cell. The variation in the productive capacities of the machines, however, can create a restriction, especially if existing machines are relocated to form a cell. The result is usually that one or two machines end up being bottlenecks, while the others are underused.

The load-balancing problem, also affected by the mix of parts entering the cell, can be alleviated somewhat by the way parts and part families are released to and sequenced through the cell. At least one company tried to attack this problem through its production planning and control system. Interestingly, however, the integration of GT cells and a scheduling system like MRP (materials requirements planning) can cause a whole new set of problems. These derive from the fact that an MRP system focuses on the completion dates of individual parts and assemblies, while cellular manufacturing focuses on the efficient production of part families. Bringing these two systems together will require new procedures for planning and scheduling.

Another problem concerns buffer inventories, which usually stack up between work stations, and the operators’ ability to eliminate these inventories in a balanced fashion. The unbalanced work load can become an advantage, however, by focusing capacity planning on these few bottlenecks. Even if the additional, dedicated machines cause a lower overall machine usage, the increased throughput times, lower inventory levels, increased productivity, and higher quality associated with cells represent a net gain. 

Two things, finally, must be pointed out regarding cellular production. First, there are, at present, no established and widely accepted formal procedures for creating cells. It is clear, however, that in the future computer simulation will increasingly be used for this purpose. Second, if cell manufacturing is implemented in an existing plant, the expectation that all parts can be allocated to part families and manufactured in production cells is unrealistic. Instead, the converted plant will be a mixture of a job shop and a cell shop, where the job shop area retains the flexibility to handle odd parts.