Ans (1)

Abduvakhidovn. (Reference Abduvakhidovna2023), Andersson and Chapma. (Reference Andersson and Chapman2017), Axtell et al. (Reference Axtell, Holman, Wall and Waterson2000), Cooper and Edget. (Reference Cooper and Edgett2010), Haddad et al. (Reference Haddad, Irene A., Hammoud, Saleh, Dwyer and Rocky2019), Hs. (Reference Hsu2008), Klei. (Reference Klein1996) and Lendel and Varmu. (Reference Lendel and Varmus2011) have explored many organisational and psychological factors and recommended technological and organisational strategies to generate innovation culture and innovation ideas. But we could not find a specific strategy to differentiate between continuous improvement in a running model and generating new products.

Ans (2)

The frameworks for product architecture range from Ulric. (Reference Ulrich1995) to Chandr. (Reference Chandra2015) and Chandr. (Reference Chandra2022). which explain the relationships between functions, components, and interfaces. But the missing link is a quantifiable parameter which one can quantitatively link to innovation in an organisation. Ulric. (Reference Ulrich1995) defines product architecture as allocation of components to functions and provides a framework for product architecture in form of three elements- functions, components, and interfaces. Though, Galvi. (Reference Galvin2001) highlights some disadvantages of modularity along with advantages and Fixson and Par. (Reference Fixson and Park2007) go even to extent of highlighting the advantages of integral architecture for innovation the wider opinion is of the view that modularity has an overwhelming advantage. Ulric. (Reference Ulrich1994), Ray and Ra. (Reference Ray and Ray2011), Galvin et al. (Reference Galvin, Burton, Bach and Rice2020), Langloi. (Reference Langlois2002), Argyres and Bigelo. (Reference Argyres and Bigelow2010), Pandremenos et al. (Reference Pandremenos, J. Paralikas, Salonitis and Chryssolouris2009), Cabigiosu et al. (Reference Cabigiosu, Francesco Zirpoli and Arnaldo Camuffo2013) and Macduffi. (Reference Macduffie2013) have characterized modularity in several ways. All of them agree that modularity helps in the process of innovation, especially in automotive domain.

As we zero down on modularity as the architectural aspect to relate to innovation, Newma. (Reference Newman2006) and Dong and Sarka. (Reference Dong and Sarkar2014) have suggested formulation to calculate the modularity of a system. But we are concerned with a specific aspect of modularity which relates to innovation. Chandr. (Reference Chandra2023) have suggested a parameter called degree of decoupling of a system which is the ratio of decoupled interfaces to the total number of interfaces in the system and established that this parameter relates to the inherent innovation potential in a product.

Ans (3)

All the approaches towards measurement of innovation e.g., Joshi et al. (Reference Joshi, Das and Mouri2015), Van De Wal et al. (Reference Van De Wal, Boone, Gilsing and Walrave2020), Diéguez et al. (Reference Diéguez-Soto, Manzaneque, González-García and Galache-Laza2019) and Jensen, and Webste. (Reference Jensen and Webster2009) look for information relating to the organization instead of a product, (whereas our objective is to relate innovation to product architecture) and they need confidential data which is difficult to get for an independent agency. Midgley and Dowlin. (Reference Midgley and Dowling2016) propose a term ‘Actualized Innovativeness’ which measures innovativeness in terms of implementation time and one can calculate it from mean and actual time of implementation and standard deviation. Ivano. (Reference Ivanov2021) also proposes a computational model for innovation, but it is based on the literature available on an idea and internet searches. So, this model is not feasible for innovations carried silently in industry. Similarly, Lhuillery et a. (Reference Lhuillery, Livramento and Raffo2015) provide a model for computing innovativeness in an organization All these methods relate to organisational factors instead of design or architectural data of a product, we need a formulation to quantify innovation in terms of innovation level and the impact of innovation.

As far as the innovation level of an innovation idea is concerned, Henderson and Clar. (Reference Henderson and Clark1990) have formulated an innovation quadrant (refer figure 1), which clearly differentiates between distinct types of innovation based on the combination of interfaces and technology.

Figure 1.

The innovation quadrant (Ref: Henderson and Clark)

Based on the Henderson and Clark framework, Chandr. (Reference Chandra2023) has suggested a quantification for the comprehensive value of innovation of an idea, which considers the level of innovation, the cost reduction impact, and the product performance improvement. One can calculate this parameter called Innovativeness Index as

(1) $$II = LCP/48$$

Where II= Innovativeness index, L= Innovation level, C= Cost reduction index, P= Impact on performance =Quantum index (QI). Category index (CI). The values of L, C, QI and PI are to be determined based on criterions provided in the study. For example, we consider the innovation idea of integrating the muffler mounting bracket (refer table 2, S. No 1 for manufacturer 1) is meant for cost reduction so P=1 and C=4 (based on the ratio of cost reduction to the product cost). Since, the innovation is architectural, we assign L=3. So, we calculate II= 3*4*1/48= 0.25. Similarly, for a product improvement idea of start stop mechanism (S. No 12 for manufacturer 1), C=1 (as this is not a cost reduction idea), CI= 4 and QI= 2 (based on the purpose of innovation and its impact on performance). Assigning L=4 (as this is a radical idea), the innovativeness index is calculated as II=4*1*4*2/48= 0.75.

4.1. Design

The most basic consideration applicable to all innovation ideas is the number of components changing. This is easily explainable as changing the design of one component means that there is no need to change any other component to interface with the changing component. But the more important consideration is that changing the design of a single component does not create a complication in servicing the product and the old design part can be simply replaced by the new design if a user wants to get his damaged or non- functional part replaced. As the number of affected parts increases, the complication in servicing multiplies as the manufacturer either must maintain the stock of old parts or replace more parts even if the user wants to replace only one part. So, the more is the number of components with change in design, the more the probability of incorporating the idea in the next model.

4.2. Product architecture

Out of the three elements of product architecture i.e. element construction, allocation of functions to elements and interfaces between the elements, one cannot generalise the first two, but the interfaces have a direct bearing on implementation strategy. If the interfaces are decoupled, i.e. if change in one component does not necessitate a change in the interfacing component, it becomes easy to implement as a continuous change. But, if the interfaces are predominantly coupled, i.e. if a change in one component necessitates a change in the interfacing component, it becomes complicated as it leads to change in the mating parts as well which multiplies the design and development effort required to implement the change. This reasoning leads us to another conclusion that the degree of decoupling of the entire system or sub- system is not relevant here but the degree of decoupling of interfaces relating to the changing components only are relevant for our consideration.

4.3. Value of innovation

There are two aspects of innovation which affect a manufacturers decision to either implement as a running change or in a new model- the category of innovation as characterized by Henderson and Clark and the value of innovation in the form of innovativeness index as suggested by Chandr. (Reference Chandra2023). Incremental innovations are those which do not result in change of element construction of function allocation or interfaces. So, they can be implemented straightaway without affecting any other part. Modular changes also result in one whole module of parts getting replaced by another module without affecting the surrounding components or interfaces. But the decision depends on the number of parts affected. Similarly for architectural changes also, if the number of affected parts is manageable from serviceability point of view, it can be implemented as a running change in existing model. But radical changes are difficult to implement as a running change as they affect many components and systems.

As far as the innovativeness index of the idea is concerned, there is no clear technological logic to justify either running change or model change. But financial, managerial, or marketing justifications can force the manufacturer to shift the innovation to a new model if the innovation value is substantial.

5.1. Step 1: formulating the decision parameter

Based on the result analysis, to meet the third objective of formulating a decision parameter, we need to define this parameter which gives a clear indication as to whether the idea is suitable to implement as a continuous improvement in the same model or the idea is to be accommodated in a new model where there is more freedom to design new parts. The result analysis clearly leads to conclude on the following element for calculating this decision parameter.

1. a) The number of changed components (n) : We have seen that the possibility of running model implementation decreases as this number increases.

2. b) The degree of decoupling (d) : A running model change becomes easier with higher degree of decoupling for the changed components.

3. c) The innovation value (ii) : This value is expressed in terms of innovativeness index of an idea and a higher value leads a manufacturer to implement in a new model in to harness greater visibility. We are not considering the innovation level separately for this parameter as this level (L) is already a part of calculation for the innovation value (ii)

So, the fundamental requirements for the parameter are following

1. 1. It should be equal to zero if there is absolutely no possibility of a continuous implementation and the value should be one when suitability for continuous implementation is maximum.

2. 2. It should increase with increase in degree of decoupling for the changed components (d) and vice- versa, the minimum and maximum values for degree of decoupling being zero and one.

3. 3. It should decrease with the value of the innovation expressed as innovativeness index (ii) and the number of changed components (n). The minimum and maximum values for ‘ii’ are zero and one and the maximum value of the parameter should be achieved only when the number of changed components is one.

We formulate a parameter called implementation index for an innovation idea to meet the requirements described above.

(2) $${\rm{I}} = (1 + {\rm{n}}_{\max } - {\rm{n}})*(1 - {\rm{ii}})*{\rm{d}}/{\rm{n}}_{\max } $$

Where I= Implementation index for an innovation idea

1. n_max= maximum number of changed components for an idea in the company

2. n = No of components to be changed due to the innovation idea

3. ii = Innovativeness index (II) of the idea

4. d= Degree of decoupling for the changed components

The flow chart for calculation of this implementation index is shown in figure 2. As we check this equation for meeting the acceptance criterions 1, 2 and 3, we get the results in affirmative. To illustrate through an example, let us consider the innovation 1 in table- 1, the two components muffler and the mounting bracket are integrated into one. Since this is an architectural innovation (linkages are changed) with L=3, C= 4 and P=1, the innovation value I is calculated to be 0.25 (ii= LCP/48). Since both the interfaces of changed components with frame are decoupled, d=1 and the number of changed components is n=2, considering n_max as 7, the implementation index comes out to be 0.643.

Figure 2.

Flow chart for calculation of Implementation Index

5.2. Step 2: collect data for innovative ideas already implemented

To collect the implemented innovative ideas, we selected two manufacturers of two- wheelers making scooters and motorcycles. The ideas are spread over many models. Some ideas are implemented as a continuous change in the same model and some ideas needed a new model to be created to implement it. We have listed some ideas in Table 2.

5.3. Step 3: classification of ideas and calculations

The next step was to classify the ideas based on architecture and innovation ladder. (a) We assigned all indices discussed earlier like cost reduction index, quantum index, category index and the innovation level index to each idea. (b) We calculated Innovativeness index for each idea by using the formula II= LCP/48. (c) We created the interface diagram for each idea, as it existed before the implementation of the idea, partially to classify the interfaces as coupled or decoupled. (d) We classified the interfaces where one can modify or change the interfacing component just by keeping the interfacing area same as decoupled and where it was not possible without altering whole of the interfacing component was classified as coupled interface as shown in table 1. (This list is not exhaustive, due to space limitation, and we have shown in this table as an example to explain the process). We have shown the interface diagram (where n₁ indicates the degree of decoupling of the changed component) for each idea (e) We calculated the degree of decoupling by dividing the number of decoupled interfaces for the changed components by total number of interfaces for the changed components. (f) Based on these steps we get the list of ideas along with Degree of Decoupling and Innovativeness Index. Please note that we have calculated the degree of decoupling of the changed components instead of that for the entire system because it is the interfaces for the changed components which are relevant for the implementation decision. We have shown the calculated values for the innovative index (i) and the degree of decoupling of the changed components (d) in table 2.

Table 1.

Example: list of innovation ideas

Table 2.

Innovation ideas and the calculation of implementation index

Note: Highlighle ideas are the ones implemented as in the same model

Legends: ii= Innovativeness index

L= Innovation level

d= Degree of decoupling tn= No of changed components II= Implementation index

A= Architectural R= Radical I= Incremental M= Modular

5.4. Step 4: result analysis

To study the exact impact of these aspects on the decision data, to meet our second objective, we plotted the number of ideas implemented in same model against those in a new model, w.r.t the number of changed components, the innovation levels, the degree of decoupling and the value of innovation for each idea (Figure 3). The histograms indicate that as long as the number of change components remains limited to one, the ideas are implemented in the same model and as this number increases the tendency to implement in the same model reduces to zero (Figure 3a). The results for innovation level (Figure 3b) show that except for radical innovations where the allocation of functions to systems themselves are changed, ideas can be implemented a continuous change for any other type of innovation. The scatter diagram for degree of decoupling (Figure 3c) sharply highlights the fact that ideas can be implemented in same model only if the degree of decoupling for the changed components is one. Otherwise, one can implement them only in a new model. On the other hand, a high value of innovation prompts the manufacturer to shift the change to a new model as it gives more visibility at the time of a new launch (Figure 3d).

Figure 3.

The effect of (a) number of changed components, (b) innovation levels, (c) degree of decoupling of changed components (d) innovation value on the implementation decision

5.5. Step 5: verification of the formulation

As we calculate the implementation index for all the ideas (table 2) for the two companies and plot the values as a scatter diagram (figure 4), the diagrams clearly and sharply delineate the ideas implemented in the same model from those implemented in a new model and we can clearly identify the threshold values. For example, for manufacturer 1, the ideas with implementation index above 0.55 are implemented in the same model. Similarly, for manufacturer 2 also, this threshold value is around 0.55. Since the points for same model and new model implementation are sharply demarcated w.r.t the threshold value, this means that the formulation fulfils the requirements, and the organisations can use it as a decision criterion.

Figure 4.

Scatter diagram for implementation index for innovation ideas and their impact on implementation decision