AI-POWERED ANALYSIS
A hybrid data-driven approach for the viable supplier selection problem: a case study of the oil and gas industry
This paper introduces a hybrid data-driven model for viable supplier selection in the oil and gas (O&G) industry. It integrates Fuzzy Best-Worst Method (FBWM), Data Envelopment Analysis (DEA), Support Vector Machine (SVM), and Random Forest (RF) to evaluate supplier performance based on sustainability, resilience, agility, and digitalization criteria. The study identifies key indicators like cost, quality, responsiveness, manufacturing flexibility, robustness, restorative capacity, pollution control, waste management, technical capability, and smart factory as most significant. Validated with an O&G case study, the model demonstrates robustness and provides managerial insights for strategic sourcing.
Executive Impact at a Glance
Deep Analysis & Enterprise Applications
Select a topic to dive deeper, then explore the specific findings from the research, rebuilt as interactive, enterprise-focused modules.
Multi-Criteria Decision Making (MCDM) & Machine Learning (ML) Integration
The research utilizes a hybrid approach combining MCDM methods (FBWM, DEA) with Machine Learning algorithms (SVM, RF) to create a robust decision-making framework for supplier selection. This integration allows for comprehensive evaluation, weighting of complex criteria, and predictive performance assessment, addressing limitations of traditional methods.
The developed SVM model achieved a remarkable 99.4% classification accuracy in predicting supplier performance, highlighting its precision for enterprise applications.
Viable Supplier Selection Process
| Method | Key Features | Consistency |
|---|---|---|
| FBWM |
|
CR values close to zero (high consistency) |
| FAHP (Traditional) |
|
Comparatively lower consistency |
Real-world O&G Industry Application
Challenge: Selecting viable suppliers in Iran's O&G sector, incorporating sustainability, resilience, agility, and digitalization aspects previously ignored.
Solution: Implemented the hybrid FBWM-DEA-SVM-RF model to evaluate 9 suppliers, identified key indicators like cost, quality, and smart factory.
Outcome: Demonstrated robustness, applicability, and validity. SVM accurately predicted ideal suppliers (100% match), outperforming expert intuition (45% error rate) and DEA alone.
Advanced ROI Calculator
Estimate your potential gains by automating supplier selection and supply chain management with AI.
Your AI Implementation Roadmap
A typical journey to integrate advanced AI for viable supplier selection into your enterprise.
Phase 1: Discovery & Strategy Alignment
Collaborative workshops to define specific supplier selection challenges, viability criteria, data availability, and strategic objectives. Outline project scope and success metrics.
Phase 2: Data Integration & Model Development
Consolidate historical supplier data. Implement FBWM for criteria weighting and DEA for efficiency labeling. Develop and train hybrid ML models (SVM/RF) using your enterprise data.
Phase 3: Validation & Pilot Deployment
Validate model performance against historical outcomes. Conduct a pilot program with a subset of suppliers, gathering feedback and refining the model for optimal accuracy and usability.
Phase 4: Full-Scale Integration & Monitoring
Integrate the AI solution into existing SCM systems. Establish continuous monitoring for supplier performance and model drift, ensuring ongoing reliability and strategic alignment.
Phase 5: Performance Optimization & Expansion
Regular reviews and updates to optimize AI models. Explore expansion opportunities to other supply chain functions, leveraging insights for broader enterprise value.
Ready to Transform Your Supplier Selection?
Leverage cutting-edge AI to build a more resilient, sustainable, and agile supply chain. Book a free consultation to discuss your unique needs.