Collaborative and integrated data-driven delay prediction and supplier selection optimization: A case study in a furniture industry

Optimizing supplier selections is an open ended problem, relevant to the operational performance of both individual companies and entire supply chains. Considering the prediction of future occurrences of delays in the optimization of supplier selections is still an under covered problem. Unlike existing literature, this article suggests a more collaborative and integrated workflow to improve the visibility and involvement of multiple stakeholders in the supplier selection decision-making processes. This is achieved through enhanced collaboration between multiple stakeholders (suppliers, customers, decision-makers from different departments, in addition to data sources from information systems), and better integration between data analysis and decision-making, through data-driven-machine-learning and optimization. The specificities of a French company in the furniture industry are considered. A workflow model is designed to support information sharing and to streamline knowledge and interactions between multiple stakeholders from different expertise domains. A Collaborative Predictive Optimization System (CPOS) is designed to classify expected occurrences of delays, to optimize order allocations, and to enable stakeholder collaboration. Delay prediction involves Decision Trees, Random Forests, and eXtreme Gradient Boosting (XGBoost). Supplier selection is solved using mathematical programming, while considering the classification of expected occurrences of delays. Stakeholder collaboration relies on information systems and uses prediction and optimization to support finding satisfactory agreements. The approach is validated using a real 3.5-year dataset, including 139 suppliers, 7,934 products and 89,080 purchase orders. A detailed experimentation, including sensitivity analysis, best–worst case analysis, and a larger scale analysis on company datasets, shows that the suggested approach enhances collaboration and achieves delay reduction and total procurement cost savings. Valuable managerial insights are collected, including the necessity to adopt digital technologies, to adapt company workflows, and to improve upstream negotiations and supplier commitments to yearly plannings.