Abstract
In the past decades specific principles were guiding product development in the space industry, like maximum performance and minimum weight. At the same time, the availability of funds driven from institutional and military budgets, as well as the existence of captive markets, made the cost of product not the critical factor in product design.
Things have now drastically changed: reduction of budgets and increase of commercial competition induced a request for minimum cost and shorter delivery time. That means that time and cost have to be taken into account as a major criterion for technical decisions.
Purpose of this paper is to present a structured “design to cost” methodology useful to manage complex product development applied to space applications. Although many attempts have been done to identify relationships between a system technical performance and the cost of its components, with the aim to better drive design with respect to market target price, these efforts have not provided more than simple tools to make preliminary estimations. An approach suitable to explicit these relationships may be more effective when based on expert judgement.
This methodology is based on the implementation of the Quality Function Deployment technique, a method which is commonly used in certain industrial sectors (i.e. automotive and consumer electronics). The QFD allows to give priority to customer's needs, to retain information, and to eliminate any ambiguity on requirement interpretation throughout company departments.
The methodology was developed and tested in Alcatel Alenia Space, the leading European company in space industry. The conventional QFD technique was modified to be adapted to the product development process implemented in the company, while the Analytic Hierarchy Process has been used to rank expert judgements and the Data Envelopment Analysis to perform selection among different design concepts.
Introduction
This paper presents a “Design to Cost” methodology to manage complex product development projects, the specific application being a satellite platform development.
In the commercial satellite market, in fact, the space industry shall be ready to recognise the changing demands coming from their customers. This requires active decision making which takes into account customers needs and willingness to pay for these needs (Wilke, 1999). But the design process is strongly driven by engineering departments where customers needs awareness is traditionally quite low.
The proposed methodology is based on the adoption of the Quality Function Deployment (QFD) in order to meet contractual requirements as well as develop a more competitive design for other potential customers. QFD, in fact, is a structured approach to define customer needs and/or requirements and translating them into specific plans to develop products meeting these needs (Schubert, 1989)
The paper is organized in four sections. The first one introduces the Target Costing and QFD concepts. The second section deals with how to make a House of Quality (HOQ). The third section illustrates the proposed methodology. The fourth section provides an application of the methodology to a platform development program for commercial telecom satellites. Conclusions sum up the most significant aspects of the method application.
Target Costing and Quality Function Deployment
Target costing is a cost management concept that has been used very effectively by leading Japanese manufacturers. It is based on a long-term, market-driven perspective rather than on a short-term, profit-driven outlook of most US manufacturing companies (Ansari, 1997). With an emphasis on market position and product leadership, target costing enables companies to attain low costs (thus ensuring a low level of prices) allowing market share retain or enlargement.
The process of Target Costing is supposed to be very interactive, and ideally most of the process activities are carried out simultaneously. In spite of this interdependence between different activities, it is useful to identify three basic phases, or levels, in the process: Market Level, Product Level and Component Level (Kato, 1995).
The first phase of the Target Costing process leads to establish the product target price. The second phase deals with the issue of meeting the target cost of the product. The heart of component-level target costing is settling the price that the firm is willing to pay for each of the externally procured components of the new product. The allocation of the overall target cost to the product components is a critical task of this process.
QFD is the method that supports this task. QFD logically begins with understanding customer needs (including the target cost) and using these needs to drive the product development process.
QFD has several definitions being used in a wide variety of industries and over a number of decades.
The American Supplier Institute defines the QFD as follows:
“A system for translating customer requirements into appropriate company requirements at each stage from research and development to engineering and manufacturing to marketing/sales and distribution. (ASI, 2001).”
The QFD methodology involves four basic phases that occur over the course of the product development process (Akao, 1997). During each phase one or more matrices, also called “Houses of Quality”, are prepared.
This QFD methodology flow is represented in Exhibit 1.
Exhibit 1 – QFD Approach
In this formal (and to some overly academic) approach, the first “House” (Product Planning) starts with an initial set of customer needs (on the left hand side of the matrix), which are then used to generate and prioritize a set of performance measures (across the top of the matrix). The most important performance measures selected from the first House are used as input of a new matrix (Assembly/Part Development Matrix). This second House is then used to generate a set of features, technologies, programs, or actions which are chosen to address the performance measures. A third matrix is used to generate manufacturing processes, and so on. Theoretically the full QFD should be used to “deploy” the Voice of the Customer throughout the product development process.
This is technically correct but quite difficult/expansive while dealing with very complex systems.
Therefore, the aim of this paper is to define a more agile, but, at the same time, effectively integrated and interdisciplinary design process suitable for systems engineering and target costing of complex systems.
This consists in focusing only on the first “House of Quality”, the Product Planning matrix.
Product Planning Matrix
The “House of Quality” matrix (see Exhibit 2) is the most recognised form of QFD (Hauser & Clausing, 1998).
It is used by a multidisciplinary team to translate a set of customer requirements, driven by market research and benchmarking data, into an appropriate number of prioritised engineering targets to be met by a new product design. There are many slightly different forms of this matrix and its ability to be adapted to the requirements of a particular problem or group of users represents its major strength.
The general format of the “House of Quality” is made up of six major components which are completed in the course of a QFD project:
1. Customer requirements - a structured list of requirements derived from customer statements.
2. Product requirements - a structured set of relevant and measurable product characteristics.
3. Interrelationship matrix - illustrates the QFD team's perceptions of interrelationships between technical and customer requirements. An appropriate scale is applied, illustrated using symbols or numbers. Filling this portion of the matrix involves discussions and consensus building within the team and may be time consuming. Concentrating on key relationships and minimising the numbers of requirements allows to reduces resource needs.
4. Competitive benchmarking - illustrates customer perceptions observed by means of market surveys. Includes relative importance of customer requirements, company and competitor performance in meeting these requirements.
5. Technical correlation (Roof) matrix - used to identify where technical requirements support or impede each other in the product design. Can highlight innovation opportunities.
6. Technical priorities, benchmarks and targets - used to record the priorities assigned to technical requirements by the matrix, to measure the technical performance achieved by competitive products and the degree of difficulty involved in developing each requirement. The final output of the matrix is a set of target values for each technical requirement to be met by the new design, which are linked back to the demands of the customer.
Exhibit 2 — The House of Quality (HOQ)
The HOQ aims to avoid that the customer needs and competitive situation are not adequately understood and taken into account.
Moreover, some of the proved benefits of adopting HOQ are reduction of time to market and improvement of customer satisfaction, as it avoids that the customer needs and many issues of the competitive environment are not adequately understood and taken into account.
A Framed Methodology for Managing a Complex Product Development
Traditional accounting and engineering processes are not sufficient while facing for today's business challenges. The necessity of an integrated and interdisciplinary cost engineering approach is always a must.
Exhibit 3 illustrates the proposed methodology used in managing a complex product development project.
Since the decisions made during the product development cycle account for 70% to 80% of product costs, their management must be initiated since the beginning of product development. Product development team shall be trained to competitive pricing or customer awareness. Target costs (accounting for a suitable profit at company level) must be established at the start and used to drive decision-making process.
If the estimated cost is greater than the target one, a cost reduction exercise shall be applied.
The cost reduction includes four macro-phases:
— Identification of Critical Components
Once the customer needs are understood and prioritized, the process continues by building up a matrix which is prepared in a very similar manner to the product planning matrix. The complex system is divided into subsystems, assemblies or parts and the relationships between them and customer requirements are established in order to provide a proper rating based on their criticality.
— Identification of Target Requirements
The process continues with a Product Planning Matrix for the most critical components. The customer needs are used as the basis for performing competitive analysis. Critical technical characteristics or product requirements are then derived. Product trade-offs are performed. Benchmarking, risks, costs and customer importance is considered in establishing product specifications. The analysis and planning done with the Product Planning Matrix becomes the basis for the critical product requirements.
— Generation of Concepts
Once product strategy, critical technical characteristics, and target values have been established with the Product Planning Matrix, product concept alternatives are developed. These concept alternatives are then evaluated using the Concept Selection Matrix.
— Selection of more effective Concept
The next step aims to select the concept more effective in order to achieve the best balance between cost, performance, and schedule.
Exhibit 3 – The Methodology Flow
“Design To Cost” in the Space Industry
The proposed methodology has been implemented by the authors in the development process of a competitive Geosynchronous telecommunication satellite platform. The critical drivers for design updating of the available platform were minimum cost and shorter delivery time.
The actual product cost was in fact higher than the target cost. Cost reduction was therefore mandatory. For this purpose a more effective QFD form was developed.
Complex systems, in fact, have many interacting requirements, design parameters and cost drivers and, as a consequence, to develop a planning matrix might an hard task. Thus, the goal of this approach was first of all to reduce the complexity to a manageable level, focusing on product “critical” components. The first step was to establish a clear mapping of all components with respect to satisfaction of the customer requirements and to their influence on the overall product cost.
Customer requirements, driven by Market Analysis, have been identified and listed in the left side of the matrix shown in Exhibit 4.
Understanding and ranking the relevance of the various requirements is a key element in target costing, as well as product development in general. In a subsequent activity customer requirements rank have been determined using the Analytical Hierarchy Process (AHP).
The AHP was developed by Saaty (1990). In the Saaty approach, a linear scale (1, 2, 3…., 9) for the pair-wise comparisons is used to quantify how much important a criterion is when compared to another one. A value of “1” for instance means “equally important”, a value of “9” means “much more important”, while at the other end a value of “1/9” means “much less important”.
As soon as these weights and main platform components were identified, an inter-relationship matrix was developed. The relationships intensity between customer requirements and the study level of detail on the product components, to be carried-on to comply with them, were assessed by using a quantitative scale. Strong correlation are identified with a “9”, moderate correlation with a “3”, and weak correlation with a “1”.
Once all customer requirements were addressed, importance ratings were calculated multiplying the customer importance rating by the weighting factor in each box of the matrix, then adding the resulting products in each column. Components ranking was obtained as result of this step.
The terminology “House of Quality” comes from the correlation matrix shown at the top of the diagram in Exhibit 2. In this matrix, the product components are compared one towards each other, in order to highlight any cross-correlation in terms of compliance with specific requirements. This correlation matrix gives the system engineer useful information to select the best way to meet all requirements, including alternative design.
Exhibit 4 shows a typical example (by an actual application) of correlation matrix, used to identify “critical” components. Moreover, costs information was added.
Exhibit 4 –Satellite Platform Planning Matrix
The general rule used for determining the “critical” zone is that improvements on more expensive components are more promising in terms of significant cost reductions.
The HoQ, as much as conceived, allows the mapping of the components on the basis of relative weight and cost. The resulting diagram represented in Exhibit 5, is divided in four areas as follows:
- Low capability of the components, of low cost, to satisfy the customer requirements;
- High capability of the components, of low cost, to satisfy the customer requirements;
- Low capability of the components, of elevated cost, to satisfy the customer requirements;
- High capability of the components, of elevated cost, to satisfy the customer requirements.
Exhibit 5 – Relative Importance Rating Diagram
As a conclusion of the analysis, the Solar Array is identified as the most critical component.
Thus, the first phase of the cost reduction process is ended and a detailed planning matrix has been developed to determine the Solar Array features (Exhibit 6).
The next step was to develop technical characteristics of the product to respond to customer needs. This was one of the most difficult steps for team members and required a good deal of facilitation and thought. The criteria for the technical characteristics were:
. Global - must not imply or constrain design alternatives to any specific technical solution or approach
. Meaningful - must be subsequently suitable to drive the design process (they can't be vague and not well defined)
. Measurable - must be able to define a target value and clearly determine whether the desired characteristic has been achieved or not
Examples of the Solar Array technical characteristics defined in order to satisfy customer needs are: mass, power, dimensions, mechanical configuration, Solar Cell efficiency, cost.
This parameters has been selected to be able to describe all Solar Array aspects with respect to design, development, manufacturing, assembly, integration on satellite, in orbit operation.
The relationship to each customer need was then established. The relationship defines the extent to which (strong, medium or weak) the actual technical characteristic complies with customer needs. After the relationships were established, the importance rating of each technical characteristic was calculated.
As technical characteristics were developed, preliminary target values were defined. This assessment is based on the previous product heritage.
In order to meet target costs and the defined technical requirements, concept alternatives were explored. A concept selection matrix with the technical characteristics as decision criteria has been used to screen these concept alternatives and to select the best concept between four alternatives obtained from various system studies. Four types of solar cells were considered. The concept selection matrix shown in Exhibit 6 lists the product requirements or technical characteristics. The technical characteristics are derived from the QFD product planning matrix. The concept alternatives are listed at the top of the matrix. Each alternative is rated according to the criterion of best matching the technical characteristics or requirements.
A non parametric multicriteria assessment methodology has been used to measure and compare concept efficiency (Norman, 1991).
Efficiency Measurement System (EMS) is a software which computes Data Envelopment Analysis (DEA) efficiency measures downloadable for academic users. Implementation of DEA formulation proposed by Banker (1984), in which input (target cost) for fixed outputs (all other) are minimised, leads to following outcomes.
Concepts considered inefficient have a relative efficiency rate lower than 100%. A concept is inefficient if a reference “virtual” concept, that can be built as a linear combination of some concept of the sample, produces at least the same amount of output performance consuming a lower input amount. Table shows that two concepts are assessed as relatively inefficient. Efficiency assessment provided by DEA shows the inefficiency degree of a concept against its reference set. However, it does not provide any ranking among the concepts.
Exhibit 6 – Concept Selection Process
Conclusions
The methodology presented in this paper is a useful approach that may be effectively implemented for modelling complex systems, the specific application being to enable an integrated and interdisciplinary design team to effectively develop a competitive satellite platform. The approach enables to compare different design solution from a technical and economic point of view. This increase the quality of the design. One of the most important factor within the process is the availability of suitable team skill and experience, as well as full transparency throughout all disciplines (enabling better team integration and system understanding).