Abstract
Traceability relationships help stakeholders understand the many associations and dependencies that exist among software artifacts created during a software development project. The extent of traceability practice is viewed as a measure of system quality and process maturity and is mandated by many standards. This paper introduces model traceability, reviews the current state of the art, and highlights open problems. One issue that impedes wide adoption of traceability is the overhead incurred in manually creating and maintaining relationships. We review the latest research advancements that address this issue through the automatic discovery of trace relationships. Model-driven development provides new opportunities for establishing and using traceability information. We discuss automatic generation of trace information through transformations and the use of traceability relationships to maintain consistency and synchronize model artifacts. We conclude with a discussion of the implementation and utilization challenges that lie ahead. © 2006 IBM.