A system dynamics framework for sense-and-respond systems

Sense-and-respond systems realize the concepts of autonomic computing at the level of business processes. One of the key requirements to build sense-and-respond systems is to accurately capture and model the dynamical behavior of business metrics, a.k.a. key performance indicators (KPI). System dynamics (SD) models and the runtime engines provide means to understand both key performance indicators and the dynamic behaviors (e.g. causality) among them. In this paper, we present a system dynamics model based upon a scenario from supply chain management domain. Our purpose is to demonstrate an alternative approach of building sense-and-respond systems. Specifically, we use system dynamics to formally define the KPIs of both the retail inventory and the supplier backlog. Additionally, we introduce objective functions and control variables as the optimization elements being part of the system dynamics formalism. Therefore, the decision (e.g. the order size from manufacturer to suppliers) would correspond to the optimal solution of the system with respect to the defined objective. These concepts will be explained through scenarios. The enabling reference architecture and deployment method using system dynamics are also presented in this paper. After the system dynamics models and corresponding components are deployed to the field, the whole system will manifest the sense-and-respond behavior in a dynamical fashion