Effect of topology on the robustness of supply networks - Metrics and results

We develop new robustness metrics for supply networks. We also propose the new ReWiSe supply network topology based on the re-wiring of the scale-free network and show that it outperforms a pure scale-free topology in some aspects when both random and targeted disruptions are likely to occur. The unique feature of our approach is that by tuning the rewiring parameter of ReWiSe it is possible to design networks with good performance on new robustness metrics in the presence of both types of disruptions. Our model is described and the experimental results and insights about choosing the right topology for achieving robustness are discussed in detail. At 20% failure rate, the ReWiSe model has higher availability and connectivity than the scale-free network by 7 to 8% and lower proximity by the same percentage. These tradeoffs are explored further.