DESIGN : Planning for Concurrent Engineering

Medical Device & Diagnostic Industry Magazine | MDDI Article Index

Originally published May 1996

Susan E. Carlson and Natasha Ter-Minassian

Among the bevy of techniques and tools that have been advanced to solve the problems faced by design and manufacturing firms, concurrent engineering has had a significant impact on the quality and development cycle of products in large corporations. Scrap was reduced by 58% and rework by 29% at McDonnell Douglas, for example, while Boeing gained a parts and materials lead time reduction of 30% and AT&T experienced a reduction in product defects of 30%.¹ It is anticipated that similar results can be attained by small companies that adopt the concurrent engineering approach to product design.

Basically, concurrent engineering is a design philosophy that seeks to improve the quality and usability of products, improve customer satisfaction, reduce costs, and ease the transition of a new product from design to manufacturing. Definitions of the term vary, but most agree that the key concepts of concurrent engineering include the use of a development team that represents all aspects of the product's life cycle, a clear understanding of customer requirements, and an early focus on production and field support systems during the design process. Although this common sense approach to design is simple in concept, its implementation can be challenging, especially to small organizations.

In addition to the support of top management, which is critical, the adoption of concurrent engineering requires a commitment of such resources as time, money, and personnel; an understanding of the process; and an ability to teach its principles to company staff and to gain their acceptance of the technique. Generally, a large company will hire a consulting firm to guide it through the implementation process. However, most small companies cannot afford this option. To assist such firms, we observed the design practices of five small design and manufacturing companies and, based on our findings, developed a methodology that explicitly describes the steps of the concurrent design process, the output of each step, and the associated team members. This article introduces that dynamic concurrent engineering methodology, which can provide small businesses with a framework for a logical design process that will increase the likelihood of producing high-quality devices that fulfill customer requirements.

CONTRASTING DESIGN STRATEGIES

The traditional sequential procedure for designing a product, shown in Figure 1, is often referred to as the "over-the-wall" approach. When this design strategy is followed, individuals are only responsible for their designated function. Marketing or upper-level management will conceive of a need and throw the idea for a new or improved device over the wall to the design engineering department. After the engineering staff has designed a product meeting the requirements given them, they will pass the project on to manufacturing, which will attempt to produce the product. However, only rarely does a new product move into production without design changes. Often, the project will move back and forth from design to manufacturing several times before the product is manufacturable within established quality levels. Then, once the product is marketed, customers will pass on their opinions of it and the process will begin again as the marketing department relays to engineering the product changes needed to accommodate customer desires. Because this method of designing does not encourage interaction between functional departments during the product development process, important issues, such as the ease of manufacture and assembly, are often left unaddressed by the design engineers.

Concurrent engineering provides an excellent alternative strategy since it encourages communication and focuses on a full identification of needs early in the product development cycle. Manufacturing and assembly requirements are expressed and considered at the same time that the needs of the end-user are investigated. As a result, fewer design iterations will be required. Although the conceptual design phase may be longer than with sequential design, because careful consideration must be given to all of the product's functional aspects, this time will be made up later in the development cycle because the product will go into full production sooner.

THE CONCURRENT DESIGN METHODOLOGY

The concurrent design methodology that is described in the remainder of this article is based on three skeletal models of a design and development process, each representing a different level of complexity. Two project assessment tools are used to determine which model or combination of models to use, as well as which personnel are most important to include on the design team.

The Project Assessment Chart. To begin the process, the project leader works with other individuals responsible for the inception of the project to complete an assessment chart, such as that shown in Figure 2. This chart tabulates the project's relative grades on each of 12 criteria: design type; product complexity; design standardization; required analytical resources; projected design cycle time; required level of precision, reliability, and durability; manufacturing process complexity; supplier requirements; size and scope of the project; project priority; risk; and cost. Each of the criteria is rated subjectively as A, B, or C based on the project manager's judgment about the company and its products and the design team's capabilities. A rating of A implies that the product needs a team that interacts often and a design process that is highly concurrent and systematic, while a rating of C implies a minimal amount of concurrency is required. The evaluation criteria are outlined in the following paragraphs.

Design Type. The type of design can vary from a completely new product, which would be rated A, to a simple feature change, such as adding a function or modifying some aspect of the user interface, which would be rated C. A B rating would be used when some of the decisions about the design have already been made because it will be based on a previous model, but substantial changes in the product's structure or function are to be made.

Product Complexity. In rating this category, the product is viewed at a micro level and its complexity is gauged based on the number and type of parts required. The number of parts is used as a measure of complexity because designing a product with many parts typically involves several design engineers who must coordinate their efforts.

For most small manufacturers, high-complexity, A-rated products are those with 50 or more parts, or with parts that will be largely designed in-house, which may present significant technological challenges for the design engineers; medium-complexity, B-rated products are those with between 20 and 50 parts, the major- ity of which will be purchased; and low-complexity, C-rated products have 20 or fewer parts, the majority of which will be purchased. It should be noted that the numbers cited are relative: for some well- established companies a highly complex product may have over 500 parts while a mid-level product might have 200. A particular product's rating must be based on the company's experience and capabilities.

Design Standardization. Either customer demands or the requirements of regulatory or other standards may restrict the potential number of design concepts that can be considered. An A rating is appropriate when there are few governing standards or specifications and many possible designs and approaches can be considered; a B rating should be given when standards or specifications may restrict parts of the design, but there is still significant room for exploring different concepts; and a C is appropriate when the number of concepts that can be considered is severely limited.

Analytical Resources. Ratings for this criterion are based on the amount of analysis required to complete the design and the analytical tools available. Such tools can include CAD systems, finite-element analysis codes, prototypes, or other methods used to verify the design before manufacturing. An A-rated product requires high levels of analytical resources, such as multiple prototypes at various stages of the design cycle and an in-depth analysis, computer aided or otherwise. A B-rated product requires fewer analytical resources, with some prototyping and computer-aided analysis, while a C-rated one needs only light prototyping and little or no computer-aided analysis.

Design Cycle Time. Long-term, medium-term, and short-term projects would be ranked A, B, and C, respectively. Again, this measure must be determined by the company, and be based on the time that the project is expected to take relative to that required by other projects in the organization.

Expected Level of Precision, Reliability, and Durability. In this category, a product would be rated A if high levels of precision, reliability, and durability are required. A product that satisfies all three of these requirements must be tested extensively under all expected operating conditions and must withstand excessive use and abuse. B and C ratings would be given if only medium or low levels of these characteristics were required.

Process Complexity. A product may be manufactured using a number of different processes, such as machining, injection molding, wave soldering, stamping, and casting. If many processes are necessary to manufacture and assemble the product, it should be rated A; if several processes are necessary to produce it, it should be rated B; and if few manufacturing processes are necessary, it should be given a C.

Supplier Requirements. This category reflects the level of supplier involvement in the product development process. When the product will include many components from outside suppliers and these parts are critical to the success of the design, a high level of supplier involvement will be needed and the product should be rated A. Similarly, a B-rated product requires that some suppliers be consulted during the design process, and a C-rated product is one that requires few or no parts or input from suppliers.

Project Size and Scope. A product design may require input from one or more engineering disciplines, such as mechanical, electrical, chemical, and software. A-rated products are those that require several engineering disciplines to participate in the design process, B-rated products require that representatives of only two disciplines work together, and C-rated ones require input from only one engineering discipline. The latter rating may include projects where the primary component of the new product will be purchased or will be designed in-house by a mechanical engineer.

Priority. The priority of a project is established on the basis of four factors: the customers' required delivery date (or the market's window of opportunity), the importance of the customers and their required level of satisfaction, the amount of capital invested in the product's development and manufacture, and the product's potential profitability. An A rating is an indication that all four factors are important, a B rating indicates that two or three factors are important, and a C rating suggests that only one of the factors is considered important.

Risk Analysis. In this assessment exercise, the term risk analysis is applied to the potential marketing life cycle of the product. For example, the manufacturer of a simple tool is likely to have a single product on the market for many years, while a computer manufacturer's product may be obsolete within six months of its release. The potential for discontinuing the project because of a volatile market should also play a role in risk analysis and priority evaluation. Low-risk, A-rated products are those that are least likely to be discontinued and are likely to have a long market life; moderate-risk, B-rated products are those where the market risk and design and development investment are in balance; and high-risk, C-rated products are those that are very likely to be discontinued and have a short-term market life.

Cost. A, B, and C ratings indicate high, medium, and low design and production costs, respectively. To accurately evaluate such costs, the project must be thoroughly researched by marketing and sales personnel, and cost estimators should be used at the outset. Cost estimation may be based on a product's prior design history or on market research.

Other issues, such as expected production volume, also help to define a product, however they do not affect the complexity of the concurrent engineering process. It is not possible to conclude that a product with a high production volume will require a more- complex design process than a one-off, because process requirements will depend on the complexity of the actual design.

The Bull's-Eye Plot. Upon completion of the project assessment chart, the results are entered on a bull's-eye graph such as that shown in Figure 3. This plot was devised to provide an indication of the optimal composition of the design team and the appropriate concurrent engineering model. It is divided into three rings, labeled A, B, and C, and 12 numbered sectors representing the categories on the assessment chart.

If most points are plotted in ring A, then model A, shown in Figure 4, is the best concurrent engineering model to use. The most complex of the three detailed models, it is intended for use by companies with the necessary personnel, technical and computing support, and financial resources to undertake a highly complex project. If the points are clustered in the ring labeled B, then model B, shown in Figure 5, should be used. This is a moderately complex concurrent engineering model, appropriate for use by companies whose resources are somewhat limited and whose products are generally variant designs of previous products. Finally, if the points are clustered in ring C, model C, shown in Figure 6, can be used. It marks the bare minimum of concurrency and is designed to meet the needs of an organization with limited resources that designs and manufactures simple products.

The scattering of points in the bull's-eye plot's 12 sections indicates which company functions the design team should be drawn from. Sections 1­5 are categories to which design engineers have the most input; sections 6 and 7 denote manufacturing input; sections 8 and 9 denote resource management and coordinator input, and include such functions as sales and marketing and financial management; and sections 10­12 cover areas where top-level managers have the most influence. If there is a cluster of points located in ring A of the design engineer quadrant, then the concurrent engineering team should be assembled with a focus on input from design engineers. Similarly, if the points cluster in the manufacturing region, input from the manufacturing engineers must be given the most weight. It is important to remember that the regions of the bull's-eye plot are not strictly delineated, and the participants' responsibilities may overlap regions.

The Design Model. The models shown in Figures 4, 5, and 6 depict the steps in three concurrent design processes of varying complexity, the participants in the steps, and the expected output from these steps. The members of the concurrent design team are identified by their initials. Those listed on the left of each design step should actively participate and provide input for that step; those listed at the right should be present at team meetings so they will be aware of decisions affecting their functions downstream in the design process. The team members and their respective development team responsibilities are listed below.

*Top-level managers (M) are responsible for maintaining a global view of company needs, including its potential profit or loss. They organize, coordinate, and oversee the financial aspects of the product lines, and are responsible for making major decisions on product development issues.

*Project managers (PM) are the core team leaders, responsible for overseeing the entire life of the design and ensuring that each phase of the development cycle is completed satisfactorily. A project manager may be an engineer or from another department of the organization, such as sales or management. In either case, he or she should have seniority and experience.

*Sales and marketing representatives (SM) are responsible for market research, identification of the customer and the customer's needs, and product sales projections. If the sales and marketing divisions are separate entities within the organization, there should be a representative from each on the development team.

*Customer representatives (C) can help identify customer needs, test prototypes, and offer criticism during design reviews.

*Design engineers (DE) design, detail, and test the product, drawing knowledge from a variety of disciplines as required. If the company has several engineering-related departments, such as mechanical, electrical, software, and test, it is useful to draw one or more individuals from each department and to co-locate them during the development project.

*Research and development staff (RD) explore new techniques for solving problems presented by functional needs in new product development that have not yet been investigated. The R&D function may be included in design engineering or may be separate, depending on the company size.

*Manufacturing engineers (ME) are responsible for all production concerns from CNC programming, machining, and assembly to system scheduling and quality assurance. They also provide input to the development team on the manufacturability of the product design.

*Resource managers (RM) are responsible for providing all of the resources and materials required in the design and production of the product. This includes locating, investigating, and coordinating component and material suppliers as well as purchasing and cost estimation. Resource management also documents past design costs and researches future design costs for purposes of accurate estimation and risk analysis.

*Financial managers (FM) have responsibility for ensuring that the financial interests of the company are considered and for setting product prices.

*Suppliers (S) are inherently linked to resource management and can provide useful input to the development team about their product lines and about customizing parts or materials to meet specific application requirements.

*Field service and repair personnel (FS) are responsible for documenting the problems associated with the company's products in their operating environment. They are an invaluable source of information on the use and misuse of a product, and on what features and functions work well, which are reliable, and which are prone to failure. Such information is particularly useful when an evolutionary redesign of a product is being carried out.

EXAMPLES

To show how this methodology can be applied to medical device design, two examples will be considered: the design of a catheter and guidewire, and the design of an infusion pump. Assume that Guide Medical, Inc., is a small company that has decided to introduce a new guided catheter into its product line because it has received feedback that physicians are unhappy with the feel of several of its current models. The company's development projects generally have a lead time of 3 to 18 months, and this design is expected to take about 3 months. Connectors and wire will be obtained from suppliers, but tubing extrusion and product assembly will be done in-house. Using this information, the bull's-eye graph shown in Figure 3 was constructed. As can be seen in this figure, most product assessment categories were rated C. However, the supplier requirements area was rated A because both the connectors and the guidewire will be purchased, and two of the three areas in the management sector were rated B. Therefore, although model C should be adopted as the guiding design model, managers and suppliers should also be included at key meetings, as shown in model B. For example, in model C suppliers and managers are not included at the design review stage, but they are included in these meetings in model B.

An infusion pump is a medium- to high-complexity device in which the mechanical, electrical, and user interfaces must be coordinated and integrated for the product to be successful. Imagine that Pumps Unlimited wants to introduce a pump model that will have two new features, one significantly different from anything the company has designed before. Because this new feature will be incorporated in all new pumps designed by the company and will give them an advantage over the competition, the design project is considered a high priority by top management. The lead time on the project is expected to be about 12 months, which falls within the usual company range of between 3 and 24 months. The cost of the pump will be approximately twice that of existing models, and process complexity will be about average, comparable with other pumps. This information was used to fill out the chart shown in Figure 2, which clearly indicates that the project should be based on concurrent design model B with no modification. Project priority was the only category rated A, so the development team must understand that this project is seen by management as important and that it should take precedence over other assigned tasks.

CONCLUSION

The implementation of concurrent engineering can be very beneficial to a company in terms of a reduction of required preproduction engineering changes, increased product quality, and the fostering of communication among all of the departments that contribute to the design and manufacture of a product. The methodology presented in this article was developed to give small companies a step-by-step process to follow that explicitly states what the outputs at each design step should be and which personnel should play a role in each of the steps. Companies that still find the technique complex and difficult are encouraged to start small and try using model C for their first attempt at concurrent design.

REFERENCE

1. Lake J, "Implementation of Multi-Disciplinary Teaming," Eng Management J, 4(2):9­13, 1992.

Susan E. Carlson, PhD, is an assistant professor in the department of mechanical, aerospace, and nuclear engineering at the University of Virginia (Charlottesville). Currently employed by IBM in Richmond, VA, Natasha Ter-Minassian was a master's degree candidate at the university when this work was completed. *