**12. Future research**

Building evolvable systems is a considerable challenge and undertaking the issues of developing evolvable systems and then formulating methods to maintain their extended life-cycles are quite large and time consuming initiatives. Given this, some future research on the selected areas of system evolution can be outlined as follows:


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

In addition, many systems will not allow suggested changes to be applied until the system is shut down for some prolonged period of time. Thus, frequent update requests will not be possible. In addition, due to time, money or manpower restrictions, many developers may not venture through setting up such a closely coupled meta-structure within the

We can think of the meta-structure attached to the system itself as a measurement program. There is a growing awareness that such systems pay off but not without some investment of both time and resources. As the benefits of applying meta-models become more evident, the establishment of such structure will become more essential for evolvable systems to retain their value to a satisfactory level and to remain competitive in the market. With the growing experiences and maturing systems, multi-dimensional data will become available for further research. This, in turn, will make it possible to develop better models. Although it is typically too costly to run carefully controlled experiments on capturing meta-data covering the entire development cycles, it is definitely worth the investment for any long-running systems. Certain meta-level system development practices must be established so that we can fully understand the system, manage it in a more cost-effective way, and readily adapt to a changing environment. Although our proposed model does not cover every aspect, it opens up a wide range of options for the system architects or stakeholders to deduce their own sets of metrics and to utilize these measures to fine-tune systems to extend their lifecycles. Eventually, the systems being instrumented to support software evolution will become more complex with higher capabilities to configure themselves with more feature-rich services. Thus, a growing proportion of resources will continue to be needed for maintenance of these systems.

Building evolvable systems is a considerable challenge and undertaking the issues of developing evolvable systems and then formulating methods to maintain their extended life-cycles are quite large and time consuming initiatives. Given this, some future research

• Automated ways to to update system objects, processes and codes needs to be

• In-depth studies of psychological impact on USI during each system upgrade will help refine the calculation of system value and focus on the influential evolutionary factors only. (Halstead's effort metrics and Cyclomatic complexity of workflow graph can be

• Applying further statistical analyses (such as vector and velocity measures for comparing TC and SV curves, applying differentiation on the system value, etc.) on the accumulated historical data in the knowledgebase will direct accurate prediction of

• Applying proper design patterns to probing stations and survey agents across the

• Because each update request needs to be addressed and implemented into the system, a corresponding queuing server model for more complex multi-project development

evolvable cycles by imposing probabilities at the state-transition points.

workflow will make the overall process more mature and reconfigurable.

on the selected areas of system evolution can be outlined as follows:

used to predict psychological complexity)

environment can be built based on queuing theories.

surrounding system environment.

**11. Conclusion** 

**12. Future research** 

investigated further

• Each system will need initialization time to configure itself and fine-tune with correct parametric values – thus a novel approach needs to be sought to generalize the initiation process with necessary delay at each the inception phase of evolution.
