To be presented at OCEANS 2018 Conference and Expedition, Charleston, SC (October 2018).

Software systems that can adapt to new requirements not only survive longer but also save overhead otherwise incurred from manually re-writing the software. Most software systems do not have this capability and older legacy software systems are frequently replaced as they become obsolete. This paper provides a two part approach for imbuing legacy systems with adaptive features. The first part utilizes program analysis, wherein we automatically transform the source code to create adaptive parameters. Second, we optimize these parameters against a wide range of operating contexts, thus expanding the programs range of capabilities beyond its original design. We demonstrate this approach on a non-adaptive case example, wherein the transformation and optimization does not rely on specific semantics of the source program. With this generality in hand, this approach may enable the expansion and optimization of a diverse variety of software systems

This material is based upon work supported by the United States Air Force and the Defense Advanced Research Projects Agency (DARPA) under Contract No. FA8750-16-C-0045. The views, opinions and/or findings expressed are those of the author and should not be interpreted as representing the official views or policies of the Department of Defense or the U.S. Government.