Evaluation of Semiconductor Risk Mitigation Strategies in the Electric Vehicle Supply Chain ^†

Abstract

The study examines the effects of semiconductor scarcity on the electric vehicle (EV) supply chain caused by an increase in electronics demand after the 2020 automobile industry downturn due to the COVID-19 pandemic. The rising demand for semiconductor chips in the automotive industry, especially in EVs, necessitates strategic measures for original equipment manufacturers and suppliers to strengthen supply chain resilience. This study uses a consequence-based decision-making framework that uses a hybrid Decision-Making Trial and Evaluation Laboratory (DEMATEL) method with Interpretive Structure Modeling (ISM). By leveraging this innovative approach, the research unveils complex causal relationships among supply chain strategies, providing quantifiable insights for prioritizing resilience in the face of multifaceted risks such as trade wars, regulatory changes, and raw material shortages. In addition, the study enhances our comprehension of supply chain resilience within the electric vehicle sector by illuminating aspects that have not been thoroughly examined by the Multi-Criteria Decision Analysis (MCDA) technique employed in this research. The analysis includes the adoption of multisourcing, fostering ecosystem partnerships, and improving supply chain visibility. Through these novel insights, this analysis aims to empower stakeholders and small- to medium-sized enterprises to navigate future automotive market dynamics, with a focus on evolving manufacturer–supplier relationships in the midst of technological advancements.

1. Introduction

Recent supply chain disruptions, including those caused by events like pandemics and natural disasters, emphasize the growing importance of strategic decisions in supply chain risk management. The COVID-19 outbreak, along with past incidents like the SARS outbreak and various natural disasters, has underscored the need for resilient supply chain networks [1]. The automotive industry, in particular, experienced a substantial 64 million global decline in sales of automobiles in the first half of 2020 due to the coronavirus [2]. China’s exports of parts decreased, Europe saw significant production cuts, and US assembly plants shut down [3]. This led to a cumulative 9.6% decline in automotive production, equivalent to approximately 7.7 million vehicles [3]. To mitigate these disruptions, Original Equipment Manufacturers (OEMs) adjusted capacity planning for critical items like semiconductors and electronic components in response to supplier challenges.

Despite a surge in demand for electronics, causing microelectronic component shortages, the automotive sector’s swift recovery is jeopardized by supply shortages, particularly semiconductors. The automotive semiconductor market is expected to grow over 9% annually until 2030, driven by increased electric vehicle (EV) adoption and advanced driver-assistance systems (ADAS) [4]. This growth, despite consistent production, exemplifies the industry’s convergence of short-term supply chain disruptions and long-term structural issues. By 2026, Battery Electric Vehicles (BEVs) are projected to dominate, with double the semiconductor content compared to Internal Combustion Engine (ICE) vehicles [4]. German manufacturers plan over 57 electric models, surpassing 100 globally. New EV registrations in Germany exceeded 100,000, increasing semiconductor demand. ADAS Level 2+ systems are anticipated to dominate, intensifying memory and logic computing demand [5]. Half of the upcoming fabrication capacity in China poses risks due to operational challenges or limited accessibility for Western companies [6].

The hybrid DEMATEL-ISM method aims to evaluate the interplay between resilient supply chain methods, determining the degree of influence and linkages [7,8] while intending to investigate the following important research questions.

(i)

What are the most important strategies that affect the resilient performance of global supply chain management during the COVID-19 outbreak?

(ii)

In the DEMATEL environment, how can the interdependencies between various resilience criteria techniques be obtained?

2. Literature Review

As per recent reports, Japan ranks third internationally and holds over 50% of the market for semiconductor material manufacturing and over 30% of the market for equipment manufacture; the US leads in providing crucial design automation software and intellectual property (IP) for chip creation. Furthermore, highlighting the value of international cooperation in developing the semiconductor sector is the US’s CHIPS and Science Act as well as Japan’s Ministry of Economy, Trade, and Industry (METI) programs [9]. Businesses are expected to prioritize building strong and resilient supply chain networks to mitigate the impact of disruptions. Key considerations include inventory planning, nearshoring, network diversification, and sourcing flexibility [6].

The literature study emphasizes nine tactics to reduce supply chain interruptions, employing the DEMATEL method to identify relationships and causal diagrams. The ISM constructs a hierarchical network model to account for dependencies and interactions across different supply chain tactics.

Table 1. Supply chain strategies.

	Strategies	Effects	References
S1	Multisourcing	Reduced dependence on a single source; less impact of supplier-specific disruptions; improved bargaining power and quality supplier	[10,11]
S2	Inventory and Capacity Buffer	Shock absorption against disruptions; time to adjust supply chains and minimize production impact	[12]
S3	Ecosystem Partnership	Collaboration and knowledge sharing; early disruption identification and risk mitigation; reduced lead times and improved responsiveness	[6,13]
S4	Platform/Plant Harmonization	Simplified production and reduced complexity; improved supplier compatibility and flexibility; faster adoption of new technologies	[14]
S5	Manufacturing Network Diversification	Reduced vulnerability to localized disruptions; lower lead times and increased demand responsiveness	[15]
S6	Nearshoring	Cut lead times and transportation costs; enhanced demand adaptability and partner collaboration; increased product development and quality	[16]
S7	Government Policy Framework	Support for industry growth and development; measures like subsidies, R&D, and supply chain diversification	[17]
S8	Supply Chain Visibility and Traceability	Reduced disruptions and improved inventory management; lower fraud and theft; enhanced product safety	[18]
S9	Vertically Integrated Supply Chain	Reduced costs, improved quality control, and time to market; increased flexibility (in some cases)	[18]

, supply chain strategies, outlines the literature review analyzed in this study, aiming to identify strategies for mitigating semiconductor risks in the electric vehicle supply chain.

3. Methodology

The Decision-Making Trial and Evaluation Laboratory (DEMATEL) is an analytical technique designed to streamline decision-making in challenging situations involving the decomposition of complex systems and the examination of key influencing elements. This method is often combined with Interpretive Structural Modeling (ISM), which breaks down intricate systems into smaller subsystems using a multi-level hierarchy. Both DEMATEL and ISM utilize relation matrices to express expert opinions on the impact connections between components [7,19]. While these techniques are commonly used independently, their integration can enhance efficiency and aid in identifying causal relationships and hierarchical patterns within complex systems. A hybrid DEMATEL-ISM technique, as proposed by [7], leverages the total-relation matrix in DEMATEL to divide the hierarchical structure of a complex system into ISM reachability matrices (Figure 1, hybrid DEMATEL ISM framework). This integrated approach is valuable for evaluating resilient solutions in the context of supply chain management. Due to the space limitations, the readers are referred to [7] for the details of the methodology.

4. Framework Implementation

The proposed framework is discussed for the resilient supply chain strategies consequence model development. In this study, nine strategies are considered based on the literature review and expert opinions. The common goal of nearshoring, multisourcing, and diversifying industrial networks is to enhance supply chain flexibility and withstand disruptions. Nearshoring, selecting suppliers and partners from the same or nearby geographic regions, saves transportation time and costs [6]. Diversification in sourcing and procurement reduces dependence on a limited number of players, enhancing adaptability to unforeseen situations. Automakers, irrespective of their markets, emphasize achieving consistency in platforms, manufacturing facilities, and products. However, expanding this approach across all product lines presents challenges, limiting its widespread use. Although organizations may view investment in inventory and capacity buffers negatively, these buffers enhance resilience and control in the supply chain, making them more robust [20].

The selection of experts for weighting criteria was based on their professional competency, task-solving experience, education level, impartiality, current position, and task-solving awareness. Experts were then given a questionnaire to rate the influence of strategies on a scale from 0 to 3, forming the Decision Matrix [21]. The Final Decision Matrix was formed by aggregating the decision matrices prepared by consultation with industry experts and also collecting data from various research papers. For example, multisourcing mitigates supply chain disruptions and enhances responsiveness to demand changes by diversifying the supplier base, reducing reliance on a single supplier, and enhancing supplier bargaining power. This, in turn, leads to improved demand forecasting, simplified manufacturing processes, reduced vulnerability to localized disruptions, and enhanced responsiveness to demand changes [6]. Inventory and capacity buffers are primarily influenced by demand forecasting accuracy and the ability to manage inventory levels. Ecosystem partnerships, platform harmonization, and manufacturing network diversification can indirectly contribute to improved inventory management by reducing disruptions, simplifying manufacturing processes, and spreading production across regions [6,13]. Ecosystem partnerships foster collaboration and knowledge sharing among supply chain stakeholders, leading to the development of innovative risk mitigation strategies, improved supply chain resilience, collaborative network planning, joint development of standardized components, and co-location of partners in specific regions [6,13]. Platform harmonization reduces component complexity, simplifies manufacturing processes, improves supplier compatibility, standardizes components and manufacturing processes across regions, and facilitates the standardization of manufacturing processes across different regions [14]. Manufacturing network diversification reduces geographical vulnerability, improves supply chain agility, expands production to new regions, and coordinates production across different regions. Nearshoring can indirectly influence multisourcing by facilitating collaborations with suppliers in different regions, enhancing supplier diversity. It influences manufacturing network diversification by expanding production to new regions and optimizing production locations for specific markets. In

Table 2. Decision matrix.

Strategies	S1	S2	S3	S4	S5	S6	S7	S8	S9
Multisourcing (S1)	0	2	3	3	3	2	1	3	0
Inventory and Capacity Buffer (S2)	0	0	1	2	3	1	0	2	1
Ecosystem Partnership (S3)	2	1	0	3	2	3	3	2	0
Plant Harmonization (S4)	1	2	2	0	1	1	0	1	1
Manufacturing Network Diversification (S5)	2	1	0	2	0	3	2	1	3
Nearshoring (S6)	2	1	1	0	1	0	2	1	1
Government Policy Frameworks (S7)	2	0	3	0	2	1	0	0	1
Supply Chain Visibility and Traceability (S8)	3	3	2	2	2	2	0	0	2
Vertical Integrated Supply Chain (S9)	0	2	1	3	2	1	2	1	0

scores are assigned as 0 for “No Influence,” 1 for “Low Influence,” 2 for “Medium Influence,” and 3 for “High Influence”.

Based on the decision matrix, the comprehensive relationship matrix is determined. After that, the degree of influence and the degree of influencing are determined in

Table 3. Causality matrix.

Strategies	Ri	Cj	Ri + Cj	Ri − Cj	Causality	Dependence Power	Driving Power
S1	3.71	2.67	6.382	1.044	cause	2	6
S2	2.21	2.61	4.823	−0.396	effect	2	1
S3	3.37	2.75	6.116	0.619	cause	2	4
S4	2.05	3.19	5.237	−1.146	effect	4	1
S5	2.89	3.36	6.247	−0.469	effect	4	2
S6	2.07	3.10	5.165	−1.027	effect	5	1
S7	2.17	2.23	4.398	−0.066	effect	1	1
S8	3.50	2.44	5.932	1.058	cause	2	6
S9	2.45	2.06	4.511	0.384	cause	1	1

, causality matrix, and Figure 2, causal diagram [8].

After developing the overall influence matrix using ISM, a threshold value is defined to generate the final reachability matrix. The driven and driving power of various strategies are determined from the final reachability matrix. The ISM methodology utilizes the reachability matrix to determine the reachability set, antecedent set, and intersection set for each factor and then assigns factors to different levels based on these sets. The hierarchical structure of the strategies is presented in Figure 3, ISM hierarchical diagraph.

5. Results and Discussions

In order to mitigate supply chain risks proactively, MICMAC analysis emphasizes the significance of driving members, such as multisourcing, ecosystem alliances, and supply chain visibility (Figure 4, MICMAC analysis). These tactics improve the overall resilience of the semiconductor supply chain by addressing the underlying causes of interruptions. In order to provide quick relief and adjust to interruptions, autonomous members like inventory buffering, vertical integration, and government support complement one another. Even if they are not directly driving or autonomous, dependent members like nearshoring, plant harmonization, and manufacturing network diversification can be increased by investing in driving and autonomous strategies. By fostering collaboration, enhancing visibility, and diversifying the supply chain, these dependent strategies can contribute to a more resilient and risk-averse semiconductor industry.

The following are the precise causal relationships between the variables:

• In regions such as East Asia, where geopolitical tensions can disrupt semiconductor supplies, reducing dependence on a single supplier through multisourcing is essential to mitigate supply chain risks;
• Ecosystem partnerships, thriving particularly in innovation hubs like Silicon Valley, Japan, and Taiwan, foster improved information exchange, creative thinking, and proactive risk reduction, ultimately strengthening the semiconductor supply chain;
• Supply chain visibility can be enhanced by implementing integrated architectures, monitoring, and forecasting requirements with collaboration across ecosystems currently critical in regions like Europe, where strict regulations mandate advanced monitoring systems to track chip movement for compliance;
• Inventory and capacity buffers can help to guarantee the continuous supply of semiconductors to automobile manufacturers;
• Nearshoring can lessen lead times, transportation expenses, and the risk of supply chain interruptions brought on by remote locations. Government policy framework can encourage local semiconductor production and reduce reliance on foreign sources, thereby strengthening the domestic semiconductor supply chain [17];
• A vertically integrated supply chain can lower reliance on outside suppliers and increase production process control, both of which can lower supply chain risk. For instance, the CHIPS and Science Act, an initiative by the US government, aims to strengthen domestic semiconductor production;
• Plant harmonization can improve responsiveness to supply chain interruptions by increasing flexibility and agility. Plant harmonization depends on resource availability and automakers’ willingness to unify technology and production procedures;
• Manufacturing network diversification can lessen the supply chain’s dependence on particular areas or suppliers while boosting its resilience. It is crucial to remember that diversifying a manufacturing network can be an intricate and expensive process [6].

Overall, (Figure 2, causal diagram) demonstrates that in order to successfully reduce supply chain risk in the semiconductor industry for the automotive industry, a combination of driving, autonomous, and dependent techniques is required. From the ISM digraph, we can find that multisourcing is at level three, which is also a main driving member. Ecosystem partnership and supply chain visibility and traceability are staged at level two, and these are also driving members as per the causal and MICMAC analysis. In comparison, inventory and capacity buffers can be linked to the supply chain visibility and traceability. Other supply chain strategies like plant harmonization, manufacturing network diversification, nearshoring, and government policy framework are interdependent at level one. However, some of them are quantified as causes in the causal diagram.

6. Conclusions

Electric car manufacturers face intense competition for market dominance, with semiconductors playing a crucial role in meeting growing demand. Supply chain disruptions in the semiconductor industry, similar to those seen during the COVID-19 pandemic, can significantly impact the electric car market’s profitability and public perception. Proactive use of supply chain mitigation techniques, especially multisourcing, is recommended based on conclusions from a hybrid Decision-Making Trial and Evaluation Laboratory Interpretive Structural Modeling (DEMATEL-ISM) approach. Businesses should keep tabs on alternate sourcing areas and reevaluate contracts with international suppliers and distribution partners on a regular basis to strengthen the resilience of their supply chains. Establishing symbiotic ecosystems with partner companies is advised for long-term sustainability, involving continuous supply and collaborative research and development efforts to stay current with semiconductor technology. Recognizing the importance of supply chain visibility, businesses are urged to use it for real-time situational awareness and informed decisions on inventory buffer stocks by utilizing technologies like digital twin and blockchain. The decision matrix, developed in collaboration with industry experts, could be further enhanced by broader expert participation and the application of Fuzzy DEMATEL and ISM methods to address higher uncertainties, improving result accuracy in uncertain situations.