A robust data driven approach to supply planning.
Abstract
We develop two robust optimization models to plan the supply operations of an assembly line when the latter are subcontracted to an external service provider. The uncertainty sets are constructed from available information on picking times both in a classical budgetbased robust approach and by using Support Vector Clustering. Numerical experiments are conducted on test instances derived from a practical case to illustrate the effectiveness of the proposed approach. The results show that the robust optimization approach is efficient to reduce the impact of picking time uncertainties on production and that the SVCbased model outperforms the classical budget-based model.
| Origin | Explicit agreement for this submission |
|---|
