Flexibility analysis in supply chain management: application to the traveling salesman problem
Résumé
During the last decades, design under uncertainty has become one of the key aspects in the Process Systems Engineering with widely spread applications to technological and economic aspects. In particular, the supply chain design is seeing a constantly increasing interest and flexibility could be an added value of critical importance for the supply chain management domain. The purpose of this research work is then to define a thorough procedure able to describe flexibility in the operational research domain taking advantage of established indicators. The most commonly used deterministic and stochastic flexibility indices are then adapted to the traveling salesman problem, an NP-hard problem in combinatorial optimization, widely studied in theoretical computer science and operations research. It can be seen as the most general problem formulation in this field and as a preliminary test before extending the flexibility assessment to more specific applications. According to existing methodologies for chemical processes, the economic and flexibility analyses have been coupled over a certain flexibility range. This approach resulted in a correlation between flexibility and costs represented by the total traveling distance. Moreover, the additional costs vs. flexibility trend can be plotted both for both the indicators showing analogies and differences with respect to the results previously obtained for chemical processes. In conclusion, the methodology provided a useful tool to quantify the parameter deviations that can be withstood keeping the solution feasible. The combined costs-flexibility graphical representation allows a more conscious decision and could be further exploited in more specific applications such as the supply chain network design.