GRAS: Non-programming approach to generate an adaptation strategy in runtime

Increasingly complex and rapidly changing software requirements have advanced the principles of adaptive software development. However, defining every potential problem scenario during the design phase that will require the system to adapt is difficult and time consuming. According to current research, defining new adaptation strategies for problems undefined during the design phase frequently requires engineers to modify the system's programs, and occasionally to modify programs in the target system. This can include scenarios when the problem is not especially critical to the system. To avoid this situation in the future, we suggest a novel approach that enables Human Actors to create an adaptation strategy through our framework, GRAS (Generator for Runtime Adaptation Strategy). GRAS, which does not require programming, provides an interface and knowledge base enabling Human Actors to participate in the adaptation process and define new strategies. Each problem scenario the system encounters is defined as a Situation to resolve ambiguity, and managed by the applicable Situation Handler, which can be created in runtime with the aid of a Human Actor. With GRAS, a Human Actor with the proper domain knowledge for a system can define additional adaptation strategies, and is not required to be askilled programmer.