Early Warning System in Agroindustry: a Chaos Theory Based Analysis

Chaos could occur anytime in agriculture sector, and that some occurences will be predictable and some will not, so it needs early warning system that could detect early chaotic conditions and take action for crisis recovery.This research was aimed to design early warning system for tapioca agroindustry with the chaos existence test. The methodology in this research we investigate the existence of chaos in agroindustry. Such an investigation is necessary to use appropriate and correct methods for further analysis, as linear system techniques will not be usefull. If a system exhibits chaos, decision making should consider the system characterization parameters from a chaos theory perspective. In this paper two models from the existing literature are reported. Of this models, the chaos existence test is further exploration to get crisis signal analysis. A tapioca smal scale industry example is used and the resulting behavior is characterized. At certain input values the behavior of the material supply system exhibits chaos. Lyapunov exponent value for the raw material supply was 0.15656 bits/week. These properties indicate that the raw material supplies could not be predicted in the long term. Raw material supply could be predicted within a period of 1/0,15 656 = 6,34942 approximately in 6 weeks. And status alert will be presented, wich is “Supply of Raw Materials have the potential Chaos”. The alert shows that tapioca industry had no strong position on integration into the upstream raw material supply sources. The recommended strategy was the strengthening of the upstream sectors This information is useful for further analysis for prediction and control. Index Term-Early Warning system, chaos theory, agroindustry