Planner based error recovery testing
A. von Mayrhauser, Michael Scheetz, Eric Dahlman, Adele E. Howe · 2002
Error recovery testing is an important part of software testing, especially for safety-critical systems. We show how an AI planning system and the concepts of mutation testing can be combined to generate error recovery tests for software. We identify a set of mutation operations on the representation that the planner uses when generating test cases. These mutations cause error recovery test cases to be generated. The paper applies these concepts to the testing of a large tape storage system.