TL;DR
This paper introduces a novel methodology for assessing and prioritizing port congestion risks specifically during the COVID-19 pandemic. By combining fuzzy Bayesian networks with AHP and a variation coefficient method, it offers a robust way to handle the inherent uncertainties in risk assessment, providing actionable insights for port authorities.
What Problem Does This Paper Address?
The COVID-19 pandemic has significantly disrupted global supply chains, leading to increased port congestion. Traditional risk assessment methods often struggle with the uncertainty and complexity introduced by such unprecedented events. Existing approaches may oversimplify risks by focusing only on likelihood and consequence, neglecting other crucial factors. This paper aims to address this gap by developing a more comprehensive and flexible risk assessment framework tailored to the unique challenges of port congestion in a pandemic environment, enabling better preparedness and mitigation strategies.
Methods
The authors propose a three-step methodology:
- Identification of Congestion Risk Factors: This involves defining port congestion risk and identifying relevant factors through literature review and analysis of congestion cases.
- A Novel Congestion Risk Model: This is the core of the paper. It builds upon Failure Mode and Effects Analysis (FMEA) by extending the Risk Prioritization Index (RPI). Crucially, it integrates a Fuzzy Belief Rule-Based Bayesian Network (FBRB-BN) to handle subjective expert judgments and model complex relationships between risk factors. To further refine the model and reduce subjectivity, the Analytic Hierarchy Process (AHP) and the variation coefficient method are combined to assign weights to different risk parameters. The model also incorporates risk utility values for calculating the final RPI.
- Validation Process: This involves an illustrative case study and a sensitivity analysis to demonstrate the model's feasibility, robustness, and effectiveness.
What's new here is the specific combination of FBRB-BN with AHP and the variation coefficient method for port congestion risk, particularly in the context of pandemic-induced uncertainty. This approach aims to overcome the limitations of traditional methods by incorporating multiple risk parameters and handling fuzzy, incomplete data.
Datasets
The paper does not explicitly mention using large-scale datasets for training or evaluation. Instead, it relies on:
- Literature Review: To identify general risk factors and understand existing research.
- Congestion Cases: Real-world examples of port congestion to inform risk identification.
- Expert Judgments: A questionnaire-based survey of five maritime experts was conducted to gather subjective probability distributions for risk parameters. These judgments were then used as input for the Bayesian network.
The study introduces a framework for risk assessment rather than a data-driven predictive model in the traditional sense.
Key Insights and Ideas
- Beyond Likelihood and Consequence: The paper emphasizes that traditional risk assessment, often limited to likelihood and consequence, is insufficient for complex scenarios like pandemic-induced port congestion. They introduce additional parameters like "Time Impact," "Additional Cost," and "Competitive Impact" to provide a more holistic view.
- Handling Uncertainty is Key: The COVID-19 pandemic introduced significant uncertainty. The use of fuzzy logic and Bayesian networks is crucial for capturing and reasoning with this uncertainty, which is often present in expert opinions and incomplete data.
- Subjectivity Needs Mitigation: While expert judgment is vital, it can be subjective. The paper's innovative combination of AHP (for subjective weighting) and the variation coefficient method (for objective weighting) aims to create a more balanced and robust weighting system.
- The "Interruption of Railways/Barges Services" is a Major Bottleneck: In the case study of the Los Angeles-Long Beach ports, this factor emerged as the most significant contributor to congestion, highlighting the critical role of inland connectivity.
Key Results
- Feasible and Functional Model: The proposed methodology was demonstrated to be feasible and functional through a case study of the Los Angeles-Long Beach (LA-LB) ports.
- Prioritization of Risk Factors: The study identified 19 risk factors and ranked them. For the LA-LB ports, the top three risk factors were:
- "Interruption of railways/barges services" (RPI: 126.899)
- "Skilled labor shortage (dockworkers or warehouse workers)" (RPI: 114.868)
- "Shortage of truck-drivers/drayage truck" (RPI: 114.209)
- Sensitivity Analysis Confirms Robustness: The sensitivity analysis indicated that the model is robust, with changes in input probabilities leading to predictable changes in risk status and RPI, supporting the model's reliability.
Limitations and Future Work
- Subjectivity in Expert Judgments: A primary limitation acknowledged is that the comprehensive weights and risk prioritization are heavily influenced by expert judgments. While efforts were made to mitigate this, subjectivity remains.
- Lack of Objective Data for Weights: The paper suggests that using objective statistical data for weight calculation would enhance accuracy and usefulness.
- Context-Specific Weights: The weights derived for the LA-LB ports may not be universally applicable and would need to be reconstructed for different ports or scenarios.
- Inter-Risk Factor Relationships: The current model does not explicitly analyze the links between internal and external risks, which could be a valuable area for future research.
- Risk Mitigation Strategies: This paper focuses on risk assessment and prioritization. Future work could explore developing risk mitigation and control strategies based on these findings.
Summary updated