Now that AOSD's first generation of languages and tools are reaching a certain level of maturity, researchers and commercial developers are adopting AOSD, primarily to develop new applications. However, there are many applications that continue to miss the advantages of AOSD, since they were written before the adoption of AOSD. Porting these applications into AOSD is currently an expensive, manual task, and the advantages, in terms of eased maintenance, are out weighed by the costs. If researchers can automate some or all of the porting process, the associated costs would drop, and porting would become worthwhile. Automation of porting can also benefit applications written using AOSD, since the organization of applications tends to deteriorate over time. Automation can help find opportunities to apply AOSD solutions that developers missed, especially in cases where parts of the scattered concern (high-level concept) span several developer's view of code. Automation of refactoring candidate identification can teach developers new to AOSD when to apply AOSD techniques, by marking concrete examples within the developer's code. Finally, the study of how to find refactoring candidates forces the research community to understand and quantify when applying AOSD solutions is beneficial.

Several interesting issues need to be addressed and studied in detail: how do we recognise crosscutting concerns in existing applications? how can we separate them from the base code? do current aspect languages suffice to express the detected concerns? how can we ensure the behavior of the existing application is preserved? how does applying AOSD affect the quality of the application? which quality attributes can be used to assess this effect?

The goal of this workshop is to address these questions, identify other interesting issues and bring together researchers from academia and people from industry working on applying AOSD techniques to already-existing applications.