Decision modeling and analysis in new product development considering supply chain uncertainties: A multi-functional expert based approach

https://doi.org/10.1016/j.eswa.2020.114016Get rights and content

Highlights

  • Bayesian theory is employed for supply chain risk components.

  • An optimization model is developed to select optimal combination of modules.

  • An example of construction power tool product line of a global firm.

Abstract

Successful new product development projects and extant research literature advocate for inclusion of inputs pertaining to the supply chain at early stages of product development to proactively identify risk averse product design concepts. To this end, we devise an analytical framework to converge upon product design concept(s) that would be associated with lesser supply chain risks, usually function of both technical and commercialization considerations. The high-level and constituent lower-level supply chain risks are represented by parent and root nodes respectively within the devised Bayesian network driven research framework. Thereafter, a quantitative measure denoted as SCRI (supply chain risk index) is evolved that yields overall composite risk numbers corresponding to respective design concepts at different risk states. Validation and comparison of the devised method with an extant study illustrates the consistency and reliability of the study. It is found that the risk propensity of a particular design concept is inversely related to the probabilistic utility of that particular concept. The case of a construction power tool of a global firm is used to demonstrate the methodology. Our research addresses an important future research pathway as argued by Hosseini et al. (2020) that extant research literature is devoid of decision-making frameworks focused on measurement and analysis the propagation of risks on complex networks.

Introduction

Conventionally, NPD (new product development) is often thought of as a stage-gate based process wherein product design concepts are selected based on broad business and technical needs, followed by realization of physical prototypes, testing and, validation, mass production, and finally product launch (Ulrich and Eppinger, 2017). However, in order to improve TTM (time to market), reduce life cycle costs, and facilitate smooth launch of new products, OEMs (original equipment manufacturers) have been increasingly taking a concurrent view of product development to include supply chain partners that are internal to the organization and external vendors, for advance and proactive validation of technical and commercial concerns (Petersen et al., 2005, Long, 2016). For instance, sometimes an OEM includes suppliers to evaluate computer aided designs (even before release of engineering drawings) of components that suppliers would be tasked to manufacture in order to gauge integrational aspects related to the product and manufacturability related concerns of components. Furthermore, with increased intricacies of contemporary manufacturing supply chains wherein several vendors supply different (with varying lead times) parts to the manufacturer for an end product, it is quite obvious that manufacturers need to consider supply chain related concerns during the design concept selection stage within NPD projects. The need to include supply chain related concerns at early stages of new product development has been also adequately emphasized by the extant research literature as well (Marsillac et al., 2014).

Within a NPD project, design concept selection is often considered to be perhaps one of the critical decisions that OEMs need to undertake for subsequent development and commercialization of associated product lines (Yang et al., 2015, Aarikka-Stenroos and Sandberg, 2012). Selecting a sub-optimal design concept may have an adverse influence on the targeted performance of enterprises in terms of escalated costs, reduced quality levels and inferior delivery performance indicators (Zhao and Cao, 2015, Zhao et al., 2014). Further, to validate design concepts, manufacturers often carryout physical testing and validation activities. These activities may include technical validations such as performance testing and market validation such as pre-launch for consumer assessment. The problem associated with such a conventional approach however is that significant resources both in monetary terms and in terms of the number of man-hours spanned across different functional agencies of the enterprise are invested. Design concept selection is driven by multifunctional inputs in terms of involvement of various supply chain stakeholders handling the product either in physical or abstract form at various stages of the value-chain (for instance, the production department for manufacturing and assembly; marketing department for gauging the mapping of product to customer needs). Thus, it would be fairly prudent to assert that design concept selection is characterized by uncertainties that, if not recognized and mitigated early in NPD properly, can derail the targeted supply chain performance of the manufacturer (Kaki et al., 2015, Zacharia and Mentzer, 2011). A key motivation for our research rationale is IBM’s design for supply chain (DfSC) program that emphasizes end-to-end collaboration to optimize a product before it takes shape in physical form (Brody and Pureswaran, 2013, Das et al., 2015). The aim here is to apply the life cycle management mentality at the design concept selection stage such that products can be developed and commercialized for optimal supply chain efficiency.

In view of the foregoing arguments, it is imperative that manufacturers evaluate feasible design alternatives in a structured manner by taking into account major supply chain concerns aimed at identification of design concept(s) associated with least overall supply chain risk. Evaluating the design concepts would however require answering the following research questions.

  • a)

    Considering that supply chain risks are often layered, how can accompanying supply chain risks be represented in various strata (s)?.

  • b)

    How can we ascertain on a composite level, overall supply chain risk index (SCRI) for a design concept that is characterized by modular architecture?

To respond to these two primary research questions, we consider a typical OEM that needs to select amongst feasible design alternatives, design concept(s) associated with least value of SCRI. Planning, sourcing, operational, logistics, market and aftersales related uncertainties are the broad components of supply chain risk considered in this research. Employing the Bayesian theory, each of the six risk components are segregated in terms of parent and root nodes. Each of these parent and root nodes are also assumed to exist in three risk states, specifically, high, medium, and low. The research problem is formulated in terms of an optimization model wherein the objective is to select the optimal combination of modules’ instance associated with the least SCRI considering constraints pertaining to operationalization of design concept, module selection, and compatibilities amongst modules’ instances. We demonstrate our research framework employing an example of construction power tool product line of a global firm. In particular, our research addresses an important future research pathway as advocated by Hosseini and Ivanov (2020) that argued that extant research literature is devoid of decision-making frameworks focused on measurement and analysis the propagation of risks on complex networks. As far as the scope of supply chain risks in the study are concerned, they are primarily regional in nature in that, the OEM primarily relies upon regional design and sourcing capabilities to ensure presence of its product line in the market space. Dimensions related to international supply chains such as risks related to sourcing from multiple international locations/off-shoring etc. are beyond the scope of our work.

The remainder of the paper is organized as follows. Section 2 presents the literature review followed by the methodology and details of the model in Section 3. 4 Model setting, 5 An illustrative case example with solution methodology enumerate the model formulation and illustrative example, respectively. The results and validation are discussed in detail in Section 6. Finally concluding remarks, limitations and pathways for future research are presented in Section 7.

Section snippets

Literature review

In this section, we discuss some of the relevant and recent research literature. In particular, our research framework evolves from primarily two domains viz. a) product design and supply chain; b) Bayesian network and probabilistic interference. We particularly consider the viewpoint from production research and operations management.

Research methodology

In our research environment, a manufacturer needs to select a particular design concept from a number of design alternatives for subsequent development and commercialization by seeking to converge at a design solution that is associated with least SCRI. Given the MCDM (multi-criteria decision-making) nature of the problem and accompanying considerations in terms of layers of risks pertaining to the supply chain, we employ a Bayesian network methodology to map risk propensities. There are two

Model setting

In our problem environment, there are V (v = 1, 2, 3……V) number of supply chain stakeholders representing the six supply chain functionalities as enumerated earlier.

The value of a supply chain risk node being judged by stakeholder v for module instance ji for module i corresponding to rth parent node being in high risk state given that tth combination of root nodes are in high risk state, can be mathematically expressed by the following term.{p(CrH/gtH)v×p(Zt,jiH)v}

Where p(CrH/gtH) is the

An illustrative case example with solution methodology

To illustrate our devised model, we present the case example of construction power tool. The respective instances for individual modules are enlisted in Table 3(b).

Referring to Table 3(b), there are a number of module instances available for selection for the design concept. These module instances are differentiated amongst themselves in terms of the manufacturing processes and material specifications. For example, the module “drill bit” contains three different module instances. These module

Results and discussions

We have considered all parent risks (that are representative of supply chain risks associated with the enterprise) to have equal weightage without one factor dominating the other. Further inputs to our developed framework are provided by stakeholders having functional expertise who have significant cross functional exposure within the organization as well as are expert in their own domain. The objective, thus, is to exploit both depth and breadth of functional knowledge (Long, 2018). The

Conclusions and limitations

The focus of our research is to integrate the decision-making pertaining to design concept selection at early stages of new product development from the standpoint of key supply chain risks such that the selected design concept represents the least supply chain risk. The supply chain risk network established is expressed in terms of risk parameters characterized by both parent and root nodes. Thereafter, employing the real-life example of construction power tool product line, we demonstrate the

CRediT authorship contribution statement

Mohit Goswami: Conceptualization, Methodology, Software, Formal analysis, Writing - original draft, Writing - review & editing. Yash Daultani: Visualization, Methodology, Writing - original draft, Writing - review & editing. Arijit De: Validation, Investigation, Writing - original draft, Writing - review & editing.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

References (38)

  • H. Shidpour et al.

    Group multi-criteria design concept evaluation using combined rough set theory and fuzzy set theory

    Expert Systems with Applications

    (2016)
  • C.S. Tang

    Perspectives in supply chain risk management

    International Journal of Production Economics

    (2006)
  • C. Tang et al.

    The power of flexibility for mitigating supply chain risks

    International Journal of Production Economics

    (2008)
  • O. Tang et al.

    Identifying risk issues and research advancements in supply chain risk management

    International Journal of Production Economics

    (2011)
  • Q. Yang et al.

    Identifying and managing coordination complexity in global product development project

    International Journal of Project Management

    (2015)
  • Y. Zhao et al.

    Risk management on joint product development with power asymmetry between supplier and manufacture

    International Journal of Project Management

    (2015)
  • Y. Zhao et al.

    An investigation of the black-box supplier integration in new product development

    Journal of Business Research

    (2014)
  • F. Alizon et al.

    Assessing and improving commonality and diversity within a product family

    Research in Engineering Design

    (2009)
  • P. Brody et al.

    The new software-defined Supply chain

    Available at

    (2013)
  • Cited by (33)

    • The adoption of new technologies for sustainable risk management in logistics planning: A sequential dynamic approach

      2022, Computers and Industrial Engineering
      Citation Excerpt :

      In their study, the model parameters were extracted through a literature review and interviews with stakeholders of the studied SC, and then the proposed model was used. Goswami et al. (2021) developed an analytical framework to converge upon the product development concepts considering SC uncertainties. In this study, a BBN was used to analyze the high-level and constituent lower-level SC risks.

    • A conceptual design decision approach by integrating rough Bayesian network and game theory under uncertain behavior selections

      2022, Expert Systems with Applications
      Citation Excerpt :

      Secondly, sub-functions are divided into core and auxiliary (i.e., non-core) types in the product design process. The core sub-function used to realize the product design requirements, while the auxiliary sub-function serves the realization of the main functions (Lu et al., 2017; Goswami et al., 2020). Compared with the auxiliary sub-function, the core sub-function has a more profound influence, and it has a greater importance in the CS evaluation process.

    • Application of blockchain for handling volatility in supply chains—a finance perspective

      2022, Blockchain in a Volatile-Uncertain-Complex-Ambiguous World
    • Application of blockchain for supply chain financing: Explaining the drivers using sem

      2021, Journal of Open Innovation: Technology, Market, and Complexity
      Citation Excerpt :

      Similarly, Ray et al. [186] tried to explore the best buyback strategies based on game theory. Again Goswami et al. [187] proposed a model for new product development amid supply chain uncertainty. The techniques presented in the above research can be embedded with blockchain to ensure better solutions for which further technical studies are essential.

    • Matrix factorization based Bayesian network embedding for efficient probabilistic inferences

      2021, Expert Systems with Applications
      Citation Excerpt :

      Finally, we repeat random sampling to generate abundant eligible samples for approximate inferences. Multiple inferences of BN could facilitate various applications, such as expert systems (Goswami, Daultani, & De, 2020), center for risk, integrity and safety engineering (C-RISE) (Khakzad, Khan, & Amyotte, 2013), medical system, and nuclear power plants, etc. For example, doctors could diagnose different patients by performing multiple inferences on the same medical diagnostic BN.

    View all citing articles on Scopus
    View full text