**4. System dynamics**

28 Real-Time Systems, Architecture, Scheduling, and Application

adjustments without the need for refactoring or overhauling the system. Thus, it is a valuable investment to the more contolled software maintenance activities. Such a wrapper system provides direct interactions among the system users, surrounding environment, and the service providers. This link (or communication path) enables the environmental responses on new changes that need to be applied as feedback into the system. Figure 1 shows an example of such placement of a wrapper system. A set of automated surveyagents and probing stations (Mubin & Luo, 2010a) are responsible for capturing various usage factors from the system itself (such as, access rates, feature ranks, utilization factors, path traversals), from its operational environment (such as, system users, work habits, workstations, workspaces, user feedbacks, new system requirements, user satisfaction

> Operational Environment **Evolvable System**

index), and from the external environment (with influential factors).

*External Environment*

Observe

System state

Capture System Usage Patterns & Feedbacks

Fig. 1. Setup of a Real-Time Meta-Structure for a System to become Evolvable

inferences, but also on the physical instant from which the outputs are produced.

The feedback concept is at the heart of the system dynamics approach (System Dynamics Society); it emphasizes a continuous view that strives to look beyond regular services to see the underlying dynamic patterns. In our setup, there are four feedback loops in the proposed operational environment through the meta-structure: (1) system usage activities are observed, and quantified into utilization factors, satisfaction indices and system state; and the knowledge base identifies whether there are any noticeable changes, and instructs the application of this change into the system, if necessary. Similarly, any change or update requests from the (2) system (3) operational environment, and (4) external influences are also applied through the service queue. Thus, over a period of time, outcomes from these activities will result in the corresponding *loop dominance*, which are then traced back into the knowledge base. The responses of such a meta-structure not only depend on logical

Activity Log & Knowledgebase

New specs

Apply changes

Apply New Changes + Generate Reports

Collect New Change Requests

New update/ change requests

> *Meta-structure (real-time system)*

**3.2 Real time feedback**

The central concept of system dynamics is the idea of two-way causation or feedback (Meadows and Robinson, 2002); it is assumed that social or individual decisions are made on the basis of the *flow* of information about the system state or environment surrounding the decision makers. The decisions lead to actions that are intended to change the state of the system. New information about the system state then produces further decisions and changes. Each such closed chain of causal relationships forms a *feedback loop*. System dynamics models are made up of such loops linked together (as shown in figure 1).

In other words, system dynamics include a set of concepts, representational techniques, and beliefs that make it into a definite modeling paradigm. It emphasizes a *continuous view* (Richardson, 1999) and looks for *causality* that underlies the longer-term patterns of change in systems. Here, the focus is on the general dynamic tendencies; under what conditions the system as a whole is stable or unstable, oscillating, growing, declining, self-correcting, or in equilibrium state (Meadows and Robinson, 2002). Properties of dynamic problems contain quantities that vary over time; such variability is described by the *causality*, which influences a closed system with *feedback loops*. Thus, dynamic systems are characterized by interdependence, mutual interaction, information feedback and circular causality. Every system experiences a number of changes over time and only in keeping detailed trackrecords of these behavioral changes can visualize the system dynamics. Therefore, the core objective is to tool the wrapper system so that we can capture dynamic system behavior, track any change history and discover correlations, as well as analyze and act upon it for better service for end users and thereby extend the system lifecycle.
