Despite the significant interest in AOSD, this emerging paradigm is still lacking strong empirical foundation. There is no testbed available for an end-to-end evaluation of AOSD techniques covering the software lifecycle phases, including requirements engineering, architecture design, detailed design, implementation, quality assurance, and evolution. Without such a testbed, AOSD techniques cannot be assessed and compared with existing ones for their contributions in terms of stringent qualities, such as modularity, reusability, and maintainability. In this context, the overall aim of this project is to develop a testbed for AOSD, called TAO, and to set up the infrastructure to capture and analyse relevant data for techniques utilising the testbed. The testbed development will be supported by the previous experience of the Lancaster AOSD team on the application of AOSD techniques to non-trivial application domains, such as multi-agent systems and databases.
The specific aims of this research project are: