Extend the meaning of "R" to "R4" in ART (automated software regression technology) to improve quality and reduce R&D and production costs

Regression testing has been conventionally employed to check the effectiveness of a solution, track existing issues and any new issues created by the result of fixing the old issues. Positioned at the tail end of the software cycle, regression testing technology can hardly influence or contribute to earlier phases such as architect, design, implementation or device testing. Extending the "R" in ART to R⁴ (regression, research, retain & grow expertise and early exposure) has been proving. R⁴ is not only providing ART with more powerful tools to detect issues as early as in the architect phase, but also arming R&D software with more proactive practices to avoid costly catastrophic problems from propagating to customer sites. This paper attempts to share some best practices and contributions from Cisco-ARF (a Cisco automated regression/research facility) whose charter is to ensure the quality of product lines running on tens of million lines of code. These award-winning practices have proven to save multi-million dollars in repair costs, thousands of engineering hours, and continue to set the higher standards for testing technology under proactive leadership and management to gain higher quality and customer satisfaction.