Network-analysis approaches to deal with causal complexity in a supply network

• Published in: International journal of production research 2012-04, Vol.50 (7), p.1840-1849
• Main Authors: Bezuidenhout, Carel Nicolaas, Bodhanya, Shamim, Sanjika, Thawani, Sibomana, Milindi, Boote, Gordon Louis Nelson
• Format: Article
• Language: English
• Subjects: Applied sciences; Causality; cause and effect; Coherence; Complex systems; Computer simulation; Decision theory. Utility theory; Dynamical systems; Dynamics; Exact sciences and technology; Flows in networks. Combinatorial problems; Graph theory; Logistics; network analyses; Network analysis; Networks; Operational research and scientific management; Operational research. Management science; Studies; Sugarcane; Supply chain management; supply network
• ISSN: 0020-7543; 1366-588X
• DOI: 10.1080/00207543.2011.575088

Large integrated supply networks can exhibit several complex system characteristics. In such systems, researchers tend to misperceive feedback relationships and have difficulty in identifying dynamic causal behaviour, even when they have an understanding of the underlying structural relationships within a system. Supply-network researchers often make use of simulation models, but this is only appropriate if a high degree of knowledge concerning the supply network is available. Many disciplines, including a limited number of supply network researchers, have used network analyses to represent complex systems, and several advanced graph theory techniques exist to support such studies. The aim of this paper is to demonstrate the use of network-analysis approaches in order to analyse supply networks. The research was carried out in four case-study areas within a sugarcane-production-and-processing environment and demonstrates two network-analysis approaches. Semi-structured interviews with stakeholders from different sectors were carried out, and issues (or problems) in the supply network were incorporated into a single coherent network. An energy-transformation approach as well as transitivity produced valuable information. A cause-and-effect network-analysis approach could depict suitable key performance indicators as well as leverage points within the supply network. These methodologies enable researchers to achieve a high degree of understanding in a relatively short time span. The analyses of the supply network occur at a higher degree of abstraction, hence obviating any need to model and understand the intricate detail of the system before any conclusion can be reached.