Developing a diagnostic heuristic for integrated sugarcane supply and processing systems.
Shongwe, Mduduzi Innocent.
MetadataShow full item record
Innovation is a valuable asset that gives supply chains a competitive edge. Moreover, the adoption of innovative research recommendations in agricultural value chains and integrated sugarcane supply and processing systems (ISSPS) in particular has been relatively slow when compared with other industries such as electronics and automotive. The slow adoption is attributed to the complex, multidimensional nature of ISSPS and the perceived lack of a holistic approach when dealing with certain issues. Most of the interventions into ISSPS often view the system as characterised by tame problems hence, the widespread application of traditional operations research approaches. Integrated sugarcane supply and processing systems are, nonetheless, also characterised by wicked problems. Interventions into such contexts should therefore, embrace tame and/or wicked issues. Systemic approaches are important and have in the past identified several system-scale opportunities within ISSPS. Such interventions are multidisciplinary and employ a range of methodologies spanning across paradigms. The large number of methodologies available, however, makes choosing the right method or a combination thereof difficult. In this context, a novel overarching diagnostic heuristic for ISSPS was developed in this research. The heuristic will be used todiagnose relatively small, but pertinent ISSPS constraints and opportunities. The heuristic includes a causal model that determines and ranks linkages between the many domains that govern integrated agricultural supply and processing systems (IASPS) viz. biophysical, collaboration, culture, economics, environment, future strategy, information sharing, political forces, and structures. Furthermore, a diagnostic toolkit based on the Technique for Order of Preference by Similarity to Ideal Solution (TOPSIS) was developed. The toolkit comprises a diagnostic criteria and a suite of systemic tools. The toolkit, in addition, determines thesuitability of each tool to diagnose any of the IASPS domains. Overall, the diagnostic criteria include accessibility, interactiveness, transparency, iterativeness, feedback, cause-and-effect logic, and time delays. The tools considered for the toolkit were current reality trees, fuzzy cognitive maps (FCMs), network analysis approaches, rich pictures (RP), stock and flow diagrams, cause and effect diagrams (CEDs), and causal loop diagrams (CLDs). Results from the causal model indicate that collaboration, structure and information sharing had a high direct leverage over the other domains as these were associated with a larger number of linkages. Collaboration and structure further provided dynamic leverage as these were also part of feedback loops. Political forces and the culture domain in contrast, provided lowleverage as these domains were only directly linked to collaboration. It was further revealed that each tool provides a different facet to complexity hence, the need for methodological pluralism. All the tools except RP could be applied, to a certain extent, across both appreciation and analysis criteria. Rich pictures do not have causal analysis capabilities viz. cause-and-effect logic, time delays and feedback. Stock and flow diagrams and CLDs conversely, met all criteria. All the diagnostic tools in the toolkit could be used across all the system domains except for FCMs. Fuzzy cognitive maps are explicitly subjective and their contribution lies outside the objective world. Caution should therefore be practiced when FCMs areapplied within the biophysical domain. The heuristic is only an aid to decision making. The decision to select a tool or a combination thereof remains with the user(s). Even though the heuristic was demonstrated at Mhlume sugarcane milling area, it is recommended that other areas be considered for future research. The heuristic itself should continuously be updated with criteria, tools and other domain dimensions.