Strategy and Design
Recall from our discussion of operations strategy that product design must be consistent with the overall strategy of the firm. In general, we would expect firms that have a low-cost strategy to have fairly standard products that change very little over time. A firm with a differentiation strategy might have a large variety of products, highly innovative products, or products with many features.
- Low Cost Strategy
- Standardized Products
- Little Change over Time
- Differentiation Strategy
- Customized Products
- Large Product Variety
- Highly Innovative Products
- Products with Many Features
It is essential for designers to know the overall strategy of the firm when designing new products. Increasingly firms are using differentiation as a way to compete. We would expect the design function to be extremely important when a differentiation strategy is employed by the firm. Designers must also be aware of basic design principles that allow new products to:
Come to market quickly,
At a competitive cost, and
Of high quality
A good design process is essential for excellent product development.
The Design Process
The stages of the design process are illustrated in the diagram below:
Generally, ideas are generated, then studied for feasibility. If the product is feasible, a preliminary design is generated from performance specifications, and a prototype is built. Process planning and final design are worked out during the next stage, and final design and manufacturing specifications are spelled out. Manufacturing begins when the product and process designs are finalized.
Idea Generation
Ideas come from many sources, including:
- Customers
- Competitors
- Suppliers
- R & D and Product Development Departments
Some companies label ideas from customers as market-pull and ideas from R & D as technology push. While ideas from customers are thought to be the most important, many products come from internal units specifically designed for product development.
Example: When IBM developed its first portable computer, individuals of exceptional talent were grouped together and sent off-site to develop the new product. Many companies use the "think-tank" approach for new products. Products that are highly technical often require development to begin with technical experts since customers have not yet thought of the product. After product introduction, customer input is essential for product improvement.
Feasibility Study
A feasibility study is an analysis to determine if a market exists for the product, and if so, will financial returns be adequate enough to support the product? Several of the analyses that must be done include:
- Market analysis -- assesses the demand for the product through demographic studies, customer surveys, interviews, focus groups, and market tests.
- Economic analysis -- examines pricing policies for the new product; compares price and cost for the sales volume, and estimates the financial return for the product
- Technical and strategic analysis -- assesses any new technology needed for the product or process; evaluates the strategic fit of the product with the rest of the firm.
Preliminary Design
A prototype is built from preliminary designs and performance specifications (instructions of what the product is supposed to do). The complete design process is an iterative process that involves building, testing, changing, testing, etc. Design is concerned with both form and function.
Form -- the physical appearance of the product
Function -- the performance of the product
Final Design and Process Planning
Final design is the stage at which detailed drawings and design specifications (product size, composition, form) are produced. Detailed drawings were called blueprints in the past, but now most detailed drawings are donewith computer-aided design (CAD). If you've ever seen blueprints or CAD drawings, you know that they contain many technical details and instructions that allow anyone working with the product to understand how it is made. Some details of process planning might also be included on the drawings. Process plans indicate which parts are purchased, which machinery must be used, how products are assembled, etc. Manufacturing specifications are detailed instructions for producing the product.
Improving the Design Process
As we mentioned earlier, more firms are using differentiation to be competitive. Since differentiation often requires uniqueness in the form of exceptional products, good design processes are essential. Poor design processes result in delayed market exposure and poor product quality. One major problem with poor design processes is excessive use of engineering change orders (ECOs). ECOs are documents that detail changes that must be made to a product after final drawings have been released. They are expensive and time-consuming.
Example: Manufactured or modular homes are becoming increasingly popular for modest-priced housing. With modular homes, framing, floors, windows, doors, and walls are constructed and assembled in a factory, then transported to a homesite. The foundations, and sometimes roofs, are made at the homesite. Suppose a window was designed into the middle of a bedroom wall, but the customer notices that he will have more room for furniture if the window is placed off-center. Changing the location of the window will require an ECO -- which means changes on the architectural drawings, rework of the existing window, scrapping of existing framing, etc. Delivery of the home may be delayed, and workmanship may suffer if workers rush to make the change.
Improving Design Processes
- Using multi-functional design teams
- Using concurrent engineering for product and process planning
- Design for manufacture (DFM) and design for assembly (DFA)
- Design for the environment
- Measuring design process quality
- Using quality function deployment (QFD)
- Designing for robustness
|
Design Teams
Design teams composed of marketing, manufacturing, and engineering are necessary for getting timely information from all relevant decision areas. Others who may be useful team members include customers, suppliers, dealers, lawyers, and accountants. Cross-functional teams are some of the most effective since each function has information unique to its own discipline. There is less danger that one person will duplicate another's work when cross-functional teams are used. Furthermore, early communication about particular functional needs allow for early planning to cover each detail.
Concurrent Design
Not so long ago, manufacturers used to talk about engineers "throwing the drawings over the wall" to manufacturing. This phrase was used in reference to the fact that engineers and manufacturers did not communicate about products very well. Engineers may have specified parts or production processes that were not available or were difficult to obtain. It was manufacturing's role to accomplish monumental feats of production without complaining. Today, we recognize that such failure to communicate is bad business. Concurrent engineering requires early manufacturing involvement or continuous manufacturing involvement.
Design for Manufacture
In recent years, more attention has been paid to technical methods of design improvement. Some important technical concepts are:
Design for manufacture (DFM) -- designing a product so it can be produced easily and economically.
Product simplification -- designing products using a minimum number of parts and simplest manufacturing processes.
Standardization -- using interchangeable parts on multiple products so that some parts can be manufactured in high volumes for economies of scale.
Modular design -- combining standardized building blocks, or modules, in a variety of ways to create unique finished products. Modular design can also mean building a few common "platforms," then adding different features to make customized products. It is a method of gaining the benefits of low cost standardized parts without losing the design flexibility that allows customers to choose their options.
Design for assembly (DFA) -- a set of procedures for reducing the number of parts in an assembly, evaluating methods for assembly, and determining an assembly sequence.
The figure below shows a good example of design for manufacture.
Example: Whirlpool is known for its excellent international products strategy. Appliances around the world differ substantially. Not only are there differences in size and aesthetic appearance, but technical requirements also differ dramatically. Electrical voltages and environmental requirements in each country make appliance design quite complicated. Yet Whirlpool was successful in finding those parts of appliances that are common across a number of countries and using them as platforms on which to add country-specific features. By using such a design strategy, Whirlpool is able to source its high volume parts at low cost from the best manufacturers in the world while using local manufacturers for its more unique parts.
Design for Environment (DFE)
Around the world (and particularly in Europe), concern for the environment is growing. New standards are being developed so that products can be designed to reduce their burden on the environment. Some design guidelines include:
- Reduce packaging
- Design for recyclability
- Design for re-use
- Design for easy repair
- Use recycled materials
- Minimize the use of materials
ISO standards for environmental design are listed under
ISO 14000.
Measures of Design Quality
When we speak about quality in this section, we are not talking about product quality. We are concerned about how well a design group has followed the principles we have discussed so far when designing a product. If a group is interested in measuring how well it is following good design principles, it should measure things like:
- Number of component parts and product options
- Percentage of standard parts
- Use of existing manufacturing processes
- Cost of first production run
- First-year cost of field service and repair
- Total product cost
- Total product sales
- Sustainable development (a measure of design for environment)
Setting good design goals and measuring them is one way to ensure that design groups will seriously pursue better ways of designing products.
Quality Function Deployment (QFD)
We've discussed the importance of communications when new products are developed. One method of guaranteeing communication among departments, suppliers, and customers is to use quality function deployment (QFD). QFD offers a structured approach for matching customer requirements with technical requirements at every stage of the design and manufacturing process. QFD uses a tool called a house of quality. The Figure below shows the parts of a house of quality.
On the left-hand side of the house, marketing would determine the essential characteristics of a product important to a customer. The "voice of the customer" drives QFD. Just below the roof, the engineers and designers would list the technical requirements of the product (related to customer needs) they typically design and specify. On the far right, the design group's product is assessed against competitors' products. The center portion of the house shows the relationship between the customer needs and technical requirements, while the roof shows how each technical requirement is related to another. The important analytical portion of the house is at the bottom. Here, the data from the other parts is analyzed for cost and technical difficulty, then rated for customer importance. Where there is a gap between what the customer sees as important and where our design group's competitive assessment is low, there should be changes made. An example of a completed house of quality is illustrated below.
The house of quality actually begins a communications chain that relates requirements for the final product to a series of requirements throughout the design and production process. Product characteristics for the final product become the customer requirements for the parts suppliers. Part characteristics then become the customer requirements for the process planners, and so on. The Figure below shows the relationships between the house of quality and successive design houses. It should be obvious that this structured approach forces a great deal of thought to be put into each level of design and production.
Design for Robustness (Taguchi's Contribution to Quality)
Robust designs are product designs that allow a product to function properly under a wide variety of environmental conditions. For example, an automobile that can as smoothly in Alaska and Arizona as it can in middleAmerica is a robust design. Conditions that cause a product to operate poorly (called noise) can be separated into controllable and uncontrollable factors. From a designer's viewpoint, controllable factors are parameters typically specified by the designer -- like material used, dimensions, and processing requirements. Uncontrollable factors are under the user's control or part of the user's environment -- things like heat, humidity, settings, degree of maintenance provided, etc. The designer's job is to choose values for the controllable factors that react to the possible occurrences of the uncontrollable factors:
Robust Design: Influence of Controllable Factors > Influence of Uncontrollable Factors
IN-LINE EXERCISE
The following factors must be considered when designing refrigerators. Indicate whether each factor is controllable or uncontrollable.
- Number of times the door is opened and shut
- Number of BTUs of thermal energy extracted by the refrigeration unit
- Number of shelves
- Amount of food stored in the refrigerator
- Size of the storage compartments
|
Reliability
Reliability -- the probability that a given part or product will perform its intended function for a specified length of time under normal conditions of use. For example, tires are expected to last for 40,000 miles when used on normal terrain. A product or system's reliability is a function of the reliabilities of its component parts and how the parts are arranged. Reliability is one of the most important measures of product quality. Reliability is often expressed as a proportion or percentage, like 90%. A reliability of 90% means that the product will operate as intended for a certain time period 90% of the time.
Let's take a look at how reliability can be used along with design for manufacture, specifically when manufacturing a system.
A system is any product composed of more than one component. Components can be connected to each other in two ways:
- In series, or in such a way that all components must operate if the product or system is to operate.
- A fuel tank -- connected to a fuel hose -- connected to a fuel pump -- connected to fuel injectors is a system with components in series.
- If one of the components in the fuel system fails, the entire system fails to operate.
- Think of a series system as water flowing through a pipe and series of valves. If one valve closes, the water flow stops.
- System reliability for series systems is expressed as
Rs = R1R2R3 . . . Rn, where n = no. of components in the system
Series Reliability
- In parallel, or in such a way that if one component fails, a similar or identical component will take over its function. Parallel systems are said to be redundant.
- The space shuttle has redundant computers. If one computers fails, another takes over its function.
- Think of a parallel system as a pipe that branches off into two pipes with valves in them, then re-joins into a single pipe on the other side of the valves. If one valve closes, water can still flow through the other side.
- System reliability for parallel systems is expressed as:
Rs = 1 - (1 - R1)(1 - R2). . . (1 - R n), or
RP=R1+R2(1-R1) if only two parallel components are present
Parallel Reliability
Another way to think about computing parallel reliabilities is to reason as in the following example:
Example: If two electrical resistors are installed in parallel, and each has a 90% reliability, that means each resistor will operate 90% of the time. One resistor may fail, and that may occur 10% of the time. During that 10% of the time, the other resistor is likely to work, but only 90% of the time. If we multiply 90% times 10%, we get 0.9, or 9% of the total time that the second resistor will provide backup. This tells us that during the time when the first resistor fails, the second resistor will work 9 out of 10 times. Since failure of the first resistor occurs only 10% of the time, our second resistor covers us 9% of the total time. To summarize, the first resistor covers us 90% of the total time, and the second covers us 9% of the total time, or 90% of the 10% failure time. Adding the reliabilities together, we get 90% for the first resistor and 9% for the second resistor, for a total of 99%. Mathematically, we can say RP=R1+R2(1-R1)
EXAMPLE
We can compute system reliabilities for different systems by reducing each component into its equivalent series reliability. Suppose we have the following system:
What is the system reliability?
The first step is to reduce the parallel reliabilities into the equivalent series reliability, so we must find the reliability of the RB - RB parallel system:
RBB = 1 - (1 - RB)(1 - RB) RBB = 1 - (1 - 0.92)(1 - 0.92) RBB = 1 - (0.08)(0.08) RBB = 0.9936
The equivalent system is now:
It is now possible to compute the system reliability since all components are at their series equivalents:
RS = RA RBBRC RS = (0.95)(0.9936)(0.99) RS = 0.9345
Suppose we did not have the redundant component at B. Our reliability would be RS = (0.95)(0.92)(0.99) = 0.8653. You can see that we improved our system reliability (0.9345) significantly by adding the redundant component B.
|
IN-LINE EXERCISE
Compute the system reliability for the system below. First compute the series equivalent for the two parallel systems, then compute the reliability for the entire system.
Compute these:
RAA=? RBB=? RCC=?
Then compute RS
|
Improving Reliability to Reduce Redundancy
Now, we see from the series reliability equation that it makes sense to keep the number of parts in a system as small as possible. As the number of parts in series increases, we increase the probability that one will fail. In some instances, however, system reliability may not be good enough, and the number of parts may have to be increased by putting them in parallel. Note the difference: in the case of parallel parts, we are installing an identical or similar part, not an entirely different part. Still, there are ways to improve reliability while also reducing redundancy:
Example: Boeing is a company that wanted to reduce the number of parts used on the Boeing 777, a wide-body jet. Its earliest wide-body jet, the Boeing 747 uses 4 engines to support the needs of the aircraft. The engines are installed in parallel so that if one engine fails, the three remaining engines will continue to operate. The Boeing 777 has only two engines -- much larger than those on the Boeing 747. Does that mean that the Boeing 777 is less reliable than the Boeing 747? No -- each individual Boeing 777 engine is made to be more reliable than each of the four engines on the Boeing 747, so the Boeing 777 is actually more reliable than the Boeing 747.
New product development is increasingly necessary to maintain competitiveness. One tool that has dramatically improved the ability to design new products is the computer.
Computer-aided design (CAD) is a software system that uses computer graphics to assist in the creation, modification, and analysis of a design. Some features of CAD include:
- geometric modeling - the use of basic lines, curves, and shapes to generate the geometry and topology of a part
- automated drafting - the production of detailed drawings directly from the CAD database
- computer-aided engineering (CAE) - simulating operating conditions of a part on the computer and analyzing it for movement, stress, interference, and thermal properties.
- link to computer-aided manufacturing (CAD/CAM) - the automatic conversion of CAD design data into processing instructions for computer-controlled equipment and the subsequent manufacture of the part as it was designed.
Benefits of CAD include:
- reduced lead time for new product introduction
- quicker product changes
- quicker product testing
- real-time access to design information
- improved design process through better communication
- more alternatives to be considered
Service Design
We live in a service economy. The proportion of services in the GDP is large, but not as large as the proportion of labor employed in services. It is more difficult to find productivity savings in services because of the large labor content. Service design, therefore must be carefully considered. Service design has some unique characteristics that differentiate it from product design. We will review these, then discuss design concepts for service.
Characteristics of Service
- Services are intangible. Services are experienced; they cannot be touched, felt, or seen. Service design is based on expectations of the experience. Many differences exist for what customers expect for the experience and from the provider.
- Services have high customer contact. Customer contact makes production of the service highly visible, unlike production of tangible goods. High visibility means that a positive encounter between the customer and provider is essential.
- Services are perishable. Since services cannot be stored in inventory, service design must define not only what is to be delivered, but also where and when.
- Consumers do not separate the service from the delivery (production) of the service. Services are produced and consumed simultaneously. Service design must specify how the service is produced, including :
- the level of customer participation in the service process
- front-room (visible) vs. backroom (invisible) tasks
- role and authority of the service provider
- the balance of touch vs. tech (level of automation)
- Services tend to be decentralized and geographically dispersed. Many small outlets make up most services. Service organizations can become large, often through franchising or "chains," but dispersion means that decision-making and control tends to be more localized.
- Services are consumed more often than products, so there are more opportunities for service "moments of truth" (chances to succeed and fail).
- Services can be easily emulated, so new service improvements may occur even more rapidly than product improvements