Designing a resilient and responsive supply chain network under disruption risks: S bi-objective stochastic programming approach

We propose a bi-objective two-stage stochastic formulation for the problem of design- ing a resilient and agile supply chain network comprising suppliers, potential locations for facilities and customers. As suppliers and facilities are vulnerable to disruption risks, various preparedness and reactive measures are considered. These include contingency procurement by switching to backup sourcing, investment in facility fortification and de- ferral of customer demand. The latter enhances the supply chain's ability to respond not only to unforeseen disruptions, but also to other sources of uncertainty, such as demand and costs. First-stage decisions define a schedule for facility deployment, the choice of fortification levels in unreliable locations and the selection of primary suppliers as well as backup suppliers. Once uncertainty is disclosed, second-stage decisions deter- mine the activation of backup suppliers and the material flows across the network. The latter may result in delayed deliveries to customers, provided that the delay does not exceed a given threshold. Two conflicting objectives are considered, namely minimising the total expected cost and minimising the total expected unmet demand. We develop a tailored two-phase heuristic procedure that is embedded in the ε-constraint method. Our numerical study with randomly generated instances demonstrates the effectiveness of the proposed methodology. Furthermore, a comparative analysis of a representative subset of Pareto-optimal solutions reveals a strong trade-off between alternative network configurations, thereby facilitating the decision-making process.