Communication Barriers to Effective Manufacturing
JAMES F. LARDNER
Empirical evidence suggests that important communication barriers exist between many of the functional groups in American manufacturing companies. There is also evidence of cultural and environmental barriers to timely, effective decision making by these organizations. These barriers prevent companies from responding rapidly and effectively to changes in market requirements and customer preferences. They also cause serious quality problems, raise product costs, and inhibit the ability of a company to compete effectively. This paper examines some of the causes of the barriers and discusses what needs to be done to overcome them.
One of the most publicized barriers is that between product design and production. In a process that has been termed “throwing-it-over-the-wall, ” design engineers concentrate on developing product functions and features with little or no consideration for how the product is to be made. Only when the design and development of the product are complete and the results are “thrown over the wall” can the production organization determine whether the product can be made at an acceptable level of quality, cost, and capital investment. The results can usually be characterized as follows:
A long time elapses between product concept and production.
The product exceeds cost goals.
Quality goals are not achieved during early production, and be-
cause production and purchasing people were not involved from the start, may never be achieved or, if so, not economically.
Excessive investment is required to produce the product at the quality and volume levels planned.
Early product life is marked by a large number of design changes—about half for reasons of quality and half for reasons of producibility.
No one accepts responsibility for failure to meet program objectives.
It is impossible to determine where responsibility for unsatisfactory results lies.
Though the relationship between design and production has received the most attention, other barriers are created by the “over-the-wall ” process. One of these relates to material sourcing and procurement, a process commonly called purchasing. With few exceptions, final material sourcing and procurement decisions can be made only when the completed product design is thrown over the wall by the design group. The results here are similar to those caused by the barriers between design and production:
Longer lead times than expected and planned are necessary to get delivery of purchased material.
Product quality objectives are not met.
Product cost targets are not met.
A large number of changes in material sourcing occur early in the life of the product.
Unnecessarily large numbers of nonstandard materials, parts, and assemblies are incorporated into the product, inflating product cost and increasing the cost of product support.
Relationships with suppliers are adversarial rather than cooperative, and it is difficult or impossible to take advantage of a supplier 's experience or technical capabilities during product design and development.
A third, less publicized, over-the-wall problem involves aftermarket product support. Service engineers are rarely involved early in product design and development. Mostly, they participate only during the final phase of product test and evaluation when they are asked to prepare product service information and recommended service parts lists for product support. Such late entry into the design and development process makes it impossible to capture critical input from service engineers based on their years of experience with past products and a great deal of knowledge about customer concerns, priorities, and use patterns. This failure results in products that are inconvenient and time consuming to service and expensive to repair.
The growth of manufacturing companies from small shops with limited
product lines employing highly skilled workers to large multiproduct enterprises with substantial concentrations of capital and labor brought a high degree of specialization and division of labor in the work force. This change created more and more narrowly focused functional groups and resulted in a significant increase in organizational complexity and in difficulty of managing the manufacturing process.
To coordinate and direct the activities of the rapidly proliferating groups of specialists, additional layers of managers were added. As this was done, management became increasingly hierarchical and autocratic. Decisions were made at the top and transmitted down to the shop floor through the functional groups because that was how management was organized. Results were reported back to the top by the same route. It became increasingly difficult to maintain effective horizontal communication and coordination at any level in the organization except at the very top.
Over time, functional groups developed their own objectives and goals and evolved separate value systems. Each group dedicated itself to the optimization of its own function with little or no regard for, or understanding of, its effect on the performance of the manufacturing whole. Performance standards and reward systems varied from group to group according to group objectives and focus.
In this environment, every interface between functions became a potential barrier to effective communication and coordination. Clearly, this is a large problem, considering the number of functions in a typical manufacturing organization, the most important of which are as follows:
Marketing and sales
Design and development engineering
Manufacturing (production) engineering
Maintenance and plant engineering
Sourcing and procurement (purchasing)
Shop floor management
Production work force
Product service (aftermarket support)
The erosion of what once had been a common manufacturing language also created barriers to communication and coordination. That erosion was caused by the developing cultural differences among functional specialties. Over time, managers of each group began to edit and selectively interpret orders coming down through the organization. It is not uncommon today to find that each functional group describes and defines the product and the processes used to make it very differently. Although this practice does not interfere seriously with vertical communication within each function, it re-
inforces the barriers to horizontal communication and coordination between the various groups at every level in manufacturing.
The disappearance of a common language among the many groups in manufacturing highlighted a previously unappreciated problem in data and information management. This is the task of translating data and information from the root sources into the format and language needed by functional groups without losing the precise intent and meaning of the original. It is apparent that there is a serious lack of discipline in the data and information management systems used by most manufacturing companies and that this lack of discipline perpetuates barriers.
There probably is little hope of reducing the number of languages used by the variety of functions in a manufacturing organization. Therefore, providing a basis for common understanding of what the root data and information mean is essential to establishing a successful manufacturing process. To provide this basis, it is critical to be able to retain the original meaning and intent of the root or seminal data and information when derivative data and information are being created by functional groups for their own use. It is evident in practice that the act of translation is the problem and that the impact of imprecise or inexact translation on the entire manufacturing process warrants much more attention than it has been given to date.
Further barriers result from careless or inadequate definitions of key terms used in manufacturing. Quality is one example. Both practical experience and the technical literature argue that quality can have a wide range of distinct aspects and that their number, nature, and relative importance vary from product to product. Thus, a scoop shovel and a 757 airplane may have some fundamental quality aspects in common, but between the products, the more detailed aspects of quality clearly differ in number, importance, and kind.
A similar problem arises when defining costs. Although aggregated costs are similar whether the product is a shovel or an airplane, the way costs are defined and measured and distributed among the various functions of each manufacturing organization varies greatly. Because specific costs are seldom properly associated with the activities that gave rise to those costs, functional groups frequently establish goals and objectives that are at cross purposes with the primary objective of the whole manufacturing organization, namely, to produce at the lowest possible cost a product that meets all the market requirements.
If the cycle of design, development, production, marketing, and product support is to be shortened and manufacturing made more efficient, all the various activities that make up manufacturing will have to be reintegrated. This implies the need to perform many of these activities concurrently rather than sequentially as they are performed today. That will demand a highly
interactive, intensively iterative environment, particularly during the development of a new product. Unless there is a data and information management system that is accessible to all participants and ensures that every participant has comparable data and information in regard to content and currency, this degree of integration will be impossible to achieve. Existing barriers will become even more damaging to the performance of the organization.
The general lack of satisfactory data and information management systems has encouraged the fractionalization of manufacturing. A manufacturing organization must react continually to changes in product requirements, product mix, product design, process design, material specifications, competitive pressures, and on and on with only brief periods of relative stability. Because of inadequate overall data and information management systems, functional groups have developed local systems in an attempt to maintain control over their own limited areas of responsibility. Since objectives and values vary from group to group, and there is little or no understanding of how the actions of one group will affect all the other groups, responses to changes in the manufacturing environment vary greatly. It is almost by accident that group actions are directed toward optimization of the whole manufacturing effort.
Manufacturing deals with continuing change, and change creates ambiguity. This is particularly acute during the process of design, development, and production. At the early stages of any program, data are scarce and subject to considerable subsequent alteration and revision. Things such as cost estimates, results of product test and evaluation, process analysis, and competitive activities continually affect design and production decisions. As the design and development process continues, however, more and better data become available and are less subject to alteration or revision. Until this begins to happen though, the absence of adequate amounts of good data inhibits decision making by project participants.
The lack of good data is a serious problem and contributes to the creation of barriers to effective project management. The problem could be minimized if there were a means of synthesizing data early in the cycle and then substituting hard data as the project progresses. The closer the synthesized data are to the ultimate “good data,” the fewer barriers there would be. This might be possible using past experience and research-based approximations. Unfortunately, there are almost no good tools to help people do this. Even though many of the decisions to be made are similar oridentical to decisions made in the past, no history of those past decisions iskept, nor are the results of those decisions analyzed and evaluated. Thus, nearly every current decision is made based on whatever may be remembered from the past and on whatever perceived results were believed to have occurred.
The uncertainty this creates is a cause of barriers to timely decision making and to achieving the best possible solutions. Capturing what has gone before and modeling the results, together with a broader ability to simulate results of certain proposed actions would not only speed up the decision-making process but produce better decisions as well.
Much decision making in manufacturing includes a high degree of uncertainty about the precise nature of the problems being addressed and about the likely result of any proposed solution. This is because of the essential nature of manufacturing itself. It is a monolithic entity, infinitely complex in all of its details but with each part so interconnected and interdependent that no part can be changed without also affecting every other part. Except for experience, there is no good way to predict the degree of change throughout the system as a consequence of a change in one of its parts. This constant uncertainty about the system response among decision makers is an important impediment to good, timely decision making.
There is a real need for a model or models to aid in determining the possible effects of various decision choices on the manufacturing whole. Development of a more accurate, better disciplined process than depending, as we do now, on the experience of a few key people and what they remember from past product programs would help greatly to improve communication and understanding among the various functional groups. It would also reduce the amount of iteration required and cut the time to arrive at acceptable solutions. With experience, it might be possible to model the interaction of increasing numbers of the various segments of manufacturing. If the effort led only to a deeper understanding of the dynamic interaction of more activities within the manufacturing whole, it would be worth doing. Time barriers to decision making are a major handicap when trying to react quickly to the market and to perceived competitive threats. A greater understanding of the possible results of any given decision could speed decision making and improve the possibility that the decision would provide a nonmalignant answer to a problem.
As the rate of change and the complexity in manufacturing increases, there is a growing need for multidisciplinary approaches to problem solving. In a product development program that extends over a comparatively long period, the number of people involved and the kinds and numbers of different disciplines and skills required vary considerably from time to time. At each point in the process, some players enter and some drop out. Others find that their status in the project team or the relative importance of their contribution changes as the project moves from stage to stage.
Unfortunately a large component of the education required for manufacturing management is experience based. Given that the traditional environment in manufacturing for the past 100 years has been one of increasing specialization and narrower focus, there has been no effective model to
train manufacturing people to view and understand the integrated manufacturing whole. There is a need today for individuals who have broad knowledge and appreciation of the total spectrum of specialties beyond their own and how those specialties fit into the manufacturing system. Few institutions that train people who enter manufacturing industries provide a comprehensive understanding of manufacturing as an integrated system. This lack of an integrated systems approach to manufacturing is a continuing barrier to better management of the process.
In summary, a considerable number of barriers to effective manufacturing are related to the inadequacy of the current data and information management system and the present data and information structures themselves. (Data and information management, in this case, means the creation, storage, transmission, transformation, derivation, and manipulation and interpretation of data and information.) Since data and information are what drive and control the manufacturing process, it is critical that the management of data and information be accorded a high priority.
There is also a serious lack of understanding by those who work in the manufacturing system of how the system works. If there really is a generic data and information structure on which the manufacturing process depends, it may be possible to guide and manage the changes in individual and functional contributions and to establish common goals and understanding by using data and information management as an tool for integration.
In a sense, data and information may be likened to electric power. It is generated and distributed throughout a system. It drives various devices (functions) that perform work within the system. It can be transformed into various forms as required to perform the work, and it must be available instantaneously throughout the system in greater or lesser intensity and quantity according to the need. As it is with electricity and power-consuming devices, without data and information to drive them, none of the multiple activities of a manufacturing system could function properly and many could not function at all. Electric power by itself has molded the physical aspect and characteristics of the manufacturing industries. More effective management of data and information could mold the intellectual component of manufacturing to create more efficient, competitive companies.