Transforming Supply Chain Decision-Making with Explainable AI

In a paper published in the journal Decision Support Systems, researchers explored how explainable artificial intelligence (XAI) could impact decision-making processes.

Study: Transforming Supply Chain Decision-Making with Explainable AI. Image Credit: Gorodenkoff/Shutterstock
Study: Transforming Supply Chain Decision-Making with Explainable AI. Image Credit: Gorodenkoff/Shutterstock

The study addressed the lag many industry players faced compared to pioneering companies in utilizing AI-driven technologies. It presented a research model tested through an experimental design with empirical data. This pioneering work provided empirical evidence on the impact of XAI on supply chain decision-making processes.

The proposed model included a serial mediation path involving transparency and agile decision-making. Findings showed that XAI enhanced transparency, significantly improving agile decision-making and cyber resilience during supply chain cyberattacks. A post hoc text analysis of tweets about XAI in decision support systems revealed a predominantly positive attitude and highlighted themes of transparency, explainability, and interpretability.

Related Work

Previous work highlights how industry leaders like Tesla have leveraged AI to optimize operations and enhance competitiveness, particularly in supply chain management. The coronavirus disease 2019 (COVID-19) pandemic accelerated the adoption of AI among companies such as Amazon and Toyota to automate and digitalize operations.

However, due to the lack of explainability, many need to catch up in utilizing AI for decision-making, especially during critical situations like cyberattacks. XAI addresses this issue by making AI processes understandable and trustworthy, thus facilitating their adoption.

Experimental Design Analysis

The study employed a vignette-based experimental design to test the hypothesized relationships. The experiment consisted of three phases: company and role scenarios, manipulated variable scenarios, and measuring variables. Following recent suggestions, news-based scenarios and a less personal approach were used to mitigate demand effects, employing a between-subjects design without identifying research questions by design.

The target population included individuals with AI and risk management experience in operations, logistics, and supply chain management. Amazon Mechanical Turk (M-Turk) was used for participant recruitment to ensure a representative sample from a diverse population.

M-Turk's database includes around 500,000 participants from 200 countries, enhancing generalizability. Best recruitment and experimental design practices were followed, utilizing a pre-screened pool of 140 individuals who met specific criteria. Participants were asked pre-screening questions about their experience with risk management and AI systems.

The experiment had three phases: familiarizing participants with a hypothetical company's supply chain manager role, manipulating independent variables (AI or XAI), and measuring responses. Participants were introduced to a scenario where they managed risks in a supply chain, focusing on mitigating cyberattacks. They were then randomly assigned to either the AI or XAI condition, reviewing scenarios with audio and images.

Variables were measured using a seven-point Likert scale, with AI transparency and agile decision-making as mediators and supply chain cyber resilience as the dependent variable. Control variables included business type and company size. AI Transparency was measured using items adapted from existing literature, assessing the AI system's ability to inform stakeholders about successes and failures.

Agile decision-making included items evaluating the availability of information, involvement in business planning, and responsiveness to environmental changes. Supply chain cyber resilience was measured by the ability to identify, cope with, and respond to cyber disruptions. The team collected participants' demographic details, ensuring a comprehensive understanding of the sample's characteristics.

Analyzing Research Findings

Regression analysis was employed to test the hypothesized relationship between research model variables and necessary tests for experimental design, method assumptions, and bias checks. The scenario test, realism test, validity test, and manipulation test confirmed the robustness of the design.

Participants' learning was strong, supporting the scenario design (mean = 5.3-5.5 out of 7). The realism check showed participants agreed the scenarios were realistic (mean = 4.08 out of 5), and the manipulation test indicated significant differences between the manipulated levels (p < 0.001). Bias checks using the marker variable technique mitigated concerns regarding common method bias (CMB).

Confirmatory factor analysis (CFA) supported the validity and reliability of the measurement model, with factor loadings above 0.7 and strong fitness indices. Regression assumptions, including normality, homoscedasticity, and multicollinearity, were satisfied. Regression models unveiled significant positive relationships among XAI, supply chain cyber resilience, AI transparency, and agile decision-making. Serial mediation analysis confirmed the significant indirect effects of XAI on supply chain cyber resilience.

Post hoc text analysis of tweets with hashtags related to XAI revealed key themes and sentiments. Data cleaning, word cloud, sentiment analysis, and topic modeling showed that XAI is seen as a tool to enhance transparency and trust in AI processes. Sentiment analysis indicated a positive attitude towards XAI, and topic modeling using the latent Dirichlet allocation (LDA)algorithm identified themes emphasizing understanding, trust, and transparency in AI systems. These findings highlight XAI's potential to improve decision support systems by making AI processes more interpretable and trustworthy.

Conclusion

To sum up, this study delved into the impact of XAI on decision-making processes, particularly within the domain of supply chain cyber resilience. Two studies employing experimental design and post hoc analysis uncovered significant positive relationships between XAI and cyber resilience, XAI and transparency, transparency and agile decision-making, and agile decision-making and cyber resilience.

The research not only advanced knowledge on XAI utilization in supply chain decision-making but also refined theoretical frameworks and offered actionable recommendations for managerial practices, emphasizing the crucial role of XAI in enhancing supply chain cyber resilience.

Journal reference:
Silpaja Chandrasekar

Written by

Silpaja Chandrasekar

Dr. Silpaja Chandrasekar has a Ph.D. in Computer Science from Anna University, Chennai. Her research expertise lies in analyzing traffic parameters under challenging environmental conditions. Additionally, she has gained valuable exposure to diverse research areas, such as detection, tracking, classification, medical image analysis, cancer cell detection, chemistry, and Hamiltonian walks.

Citations

Please use one of the following formats to cite this article in your essay, paper or report:

  • APA

    Chandrasekar, Silpaja. (2024, June 07). Transforming Supply Chain Decision-Making with Explainable AI. AZoAi. Retrieved on December 10, 2024 from https://www.azoai.com/news/20240607/Transforming-Supply-Chain-Decision-Making-with-Explainable-AI.aspx.

  • MLA

    Chandrasekar, Silpaja. "Transforming Supply Chain Decision-Making with Explainable AI". AZoAi. 10 December 2024. <https://www.azoai.com/news/20240607/Transforming-Supply-Chain-Decision-Making-with-Explainable-AI.aspx>.

  • Chicago

    Chandrasekar, Silpaja. "Transforming Supply Chain Decision-Making with Explainable AI". AZoAi. https://www.azoai.com/news/20240607/Transforming-Supply-Chain-Decision-Making-with-Explainable-AI.aspx. (accessed December 10, 2024).

  • Harvard

    Chandrasekar, Silpaja. 2024. Transforming Supply Chain Decision-Making with Explainable AI. AZoAi, viewed 10 December 2024, https://www.azoai.com/news/20240607/Transforming-Supply-Chain-Decision-Making-with-Explainable-AI.aspx.

Comments

The opinions expressed here are the views of the writer and do not necessarily reflect the views and opinions of AZoAi.
Post a new comment
Post

While we only use edited and approved content for Azthena answers, it may on occasions provide incorrect responses. Please confirm any data provided with the related suppliers or authors. We do not provide medical advice, if you search for medical information you must always consult a medical professional before acting on any information provided.

Your questions, but not your email details will be shared with OpenAI and retained for 30 days in accordance with their privacy principles.

Please do not ask questions that use sensitive or confidential information.

Read the full Terms & Conditions.

You might also like...
Enhancing AI Transparency: Researchers Develop Advanced Watermarking Techniques