Exploring the Causal Model of 3D-CAD Technology: Interfirm Communication and Product Development Performance in Japanese Automobile Parts Industry

: 3D-CAD technology impacts the performance of product development in two ways. One is a positive effect that the use of the 3D-CAD decreases the number of problem solving cycles directly, so that development performance improves. The other is the negative effect that 3D-CAD increases interfirm communication, which increases the quantity of coordination. In conclusion, the overall phenomenon of a correlation between the use of 3D-CAD technology and improved performance cannot be seen, because the two effects of the 3D-CAD offset each other.


Introduction
Over the past few years, many existing studies on 3D-CAD information technology have examined what kinds of influence 3D-CAD has had on product development and its process (e.g., Aoshima, 1998;Aoshima, Nobeoka & Takeda, 2001;Takeda 2000aTakeda , 2000bNobeoka, 1997;Robertson & Allen, 1993;). Many empirical studies including  have reported that the introduction and use of 3D-CAD has the effect of 'improvement of product quality.' However, they have reported that the use of 3D-CAD is irrelevant to 'the development efficiency (cost and development lead time).' In other words, there were inconsistencies between the theoretical expectation effect and the empirical study results of 3D-CAD technology.
In the present study, we were not able to find the relations (correlation) between 3D-CAD technical use and development efficiency when we If a conclusion is taken in advance, it was found that 3D-CAD technology has effects on the development efficiency through two paths. That is, one path is that 3D-CAD technology reduced the number of problem solving cycles and it contributed to shortening of development lead time. The other path is that interfirm communication increased and inversely gave the influence that seemed to let the number of problem solving cycles increase. In other words, because a negative effect offsets a positive effect, correlation between 3D-CAD use and the development efficiency was not found.
This result indicated that company should not merely require the efficiency of 3D-CAD as a communication tool. That is to say, it may imply that engineers of a division and/or suppliers can participate in design communication process more actively and frequently by a computer screen, and this can largely improve product quality.
Actually, 3D-CAD can provide high visibility and enable engineers to check easily many parts interference problems on a screen and do simulations such as a check or a function test of part interference with lower cost and enable digital virtual assembling (e.g., Aoshima, 1998;Baba & Nobeoka, 1998;Nobeoka, 1997;Robertson & Allen, 1993;Takeda, 2000b;Takeda, Aoshima, & Nobeoka, 2000). In this point, the result of this paper indicates that 3D-CAD improves quality of development performance rather than improving development speed directly.

Communication
The typical good example that revolutionized a product development process with 3D-CAD technology is the development of Boeing 777. The development lead time of Boeing 777 was greatly shortened by the use of 3D-CAD technology. It became easy to ascertain the problem of many parts interference on a computer screen by introducing 3D-CAD technology into the design process of the aircraft which is an extremely complicated product.
As a result, a check and a function test of component interactions were enabled at an early stage and at low cost, which brought rapid improvement of development performance (e.g., Aoshima, 1998;

3
Baba & Nobeoka, 1998). In other words, with 3D-CAD, digital virtual assembling can be done on a screen in the middle of a design process where detailed product information is converted into digital data.
By fully utilizing its simulation function, the front loading of problem solving became conventionally easy, and it directly improved design quality.
Simultaneously, by reducing the number of prototypes, 3D-CAD had effects on shortening development lead time and reducing development cost (Aoshima, 1998;Nobeoka, 1997;Thomke & Fujimoto, 2000).  and Nobeoka, Takeda and Aoshima (2002) who studied machinery-related manufacturing companies report that increase of use rate of 3D-CAD brings improvement of product quality much as many studies have reported until now. In addition, their studies point out that the deficiency of organizational actives for a task-redefining occurrs by introducing 3D-CAD technology, which has a negative effect on development efficiency in the short term.
On the other hand, 3D-CAD, being able to provide three dimensional image of product information, can translate tacit knowledge of product information to explicit knowledge. Therefore, designers can have nearly the same image of a product and the formation of information. In this sense, 3D-CAD technology has a large effect on communication, which is regarded as an important determinant factor in the efficiency of concurrent engineering. Robertson and Allen (1993), Baba and Nobeoka (1998), and Takeda (2000a)  The automobile, which is the object of this study is an integral architecture product, that is, interdependences between parts are high and complex. The pattern of product development tends to be an overlapping type linking design stage and engineering stage, because functional and structural adjustments between parts are frequently required design changes by the auto maker, for the sake of cooperation and coordination, communication between auto maker and suppliers is extremely important in the design stage (Clark & Fujimoto, 1991;Dyer, 1996;Fujimoto, 1997Fujimoto, , 1998.
The introduction of 3D-CAD technology to the auto industry greatly contributed to shorten development lead time of Japanese auto makers.
Japanese auto makers use 3D-CAD and CAE (Computer-Aided Engineering) technology, which allowed them cost reduction and shortening of Ku 4 development lead time by making a prototype of high completeness at an early stage (Nobeoka, 1997;Fujimoto, Nobeoka, Aoshima, Takeda, & Oh, 2002).
To shorten vehicle development lead time, it is necessary for a supplier to shorten component development lead time, prototype building, production, and delivery.
This implies that suppliers must build the system which is regularly connected with auto maker's system. Also, as a supplier,

Research Method and Sample
The data were collected in a questionnaire survey called 'Questionnaire survey 1999 for auto part business in the auto industry' (Fujimoto, Matsuo & Takeishi, 1999) and interviews.

Causal Model of 3D-CAD Technology
5 Therefore we will build a causal model to explain such a phenomenon.

Building Causal Model
The points that have become clear by existing studies are summarized as follows.

Covariance Structure Analysis (Structural Equation Modeling)
This paper conducts covariance structure analysis. Table 2 and Table 3 show observed variables to constitute each latent variable and an average and variance of these observed variables.
The following result was obtained (Figure 2).
At first, the Chi-square value of the model is 29.724 and p value is 0.157. Here, Chi-square statistic indicates whether the model has covariance structure.
Let us see the goodness of fit of this model.
As p value is 0.157 and so exceed 0.05, null hypothesis that the model has covariance structure is not rejected. Thus, the causal model has a good fit (GFI = 0.952, AGFI = 0.906). Therefore, we can accept this causal model. Interpretation follows below.
As all of the causal coefficients between latent variables and observed variables used in the model

Causal Model of 3D-CAD Technology
9 are significant at the 5% level, we can confirm that all constructs were adequately measured by these observed variables. Moreover, the paths among all latent variables were also statistically significant at the 5% level.

Conclusion and Discussions
This paper focused on why the correlation between the introduction and use of 3D-CAD technology and the development efficiency has not been observed.
Considering the quantity of communication contributes to the improvement of quality and function of a product.
As we mentioned before, with 3D-CAD it is possible to transfer the shape, structure and attributes of a product via digital data. The skills and  Particularly, in recent years, the exchange of drawing information and design changes are done by 3D data. As a supplier, the introduction of 3D-CAD is a basic condition in transaction. According to interviews, because there are many cases that even participation is not allowed in development competition unless a supplier is equipped with 3D-CAD systems with compatibility. In other words, suppliers must introduce a system which is compatibile to the auto maker's system to maintain business relations. In this point, it seems that the introduction of 3D-CAD in suppliers is often a passive act from business needs rather than from strategic needs. It is assumed that such a passive introduction of 3D-CAD causes imbalance in the skill and degree of understanding between auto maker and supplier, and thereby increases unnecessary communication.
Secondly, as have been pointed out, it can be supposed that the finding is caused by over-coordination and over-examination. As a result, The problem in the auto parts industry is quite similar to 'the paradox of 3D-CAD introduction' pointed out by , and Nobeoka, Takeda, and Aoshima (2002). In other words, it is thought that the finding is caused by the fact that short use period of 3D-CAD technology, the increase of unnecessary man-hours caused by the imbalance between organizations for a newly technical knowledge system, and an organization's internal resistance.
In this sense, it may be said that the survey period examined by this paper is a transition period, in which the 3D-CAD technology is undergoing introduction processes.