Practical Process Non-Conformance Detection
Sean M. Thompson, Torab Torabi · Advances in Computational Sciences and Technology · 2011
This paper represents the culmination of a thorough investigation into the existing technological methodologies to solve process non-conformance detection. We explore how to develop this knowledge into an approach that offers a useable, practical solution. Process non-conformance detection research has been explored fairly sparingly and sporadically from around 1995 until now. The work presented in the literature, while technically effective at detecting non-conformance seemed to be practically unviable for a variety of reasons. In this paper we address why the detection of process nonconformance is valuable for process improvement, we examine the practical shortcomings in the existing solutions, explain why these shortcomings exist and how these approaches have failed to provide practical value in various ways. We explore the issues currently related to this research area that withhold process non-conformance detection from being practically useful, and present our own approach that addresses these issues. Our approach involves a generic and practical method of specifying rules that describe how enactment data is to be checked for any non-conformance instances that may have occurred. The approach is designed to cater for practical considerations such as timely detection, minimal interference, and minimal required effort while remaining separate from the process specification. We support this claim via two case studies that test our assertions and demonstrate how process non-conformance detection is valuable for process improvement. Furthermore, we show how a process may be improved by the approach using the result data from several hundred process enactments done both before and after the approach was integrated. The result data from these two phases is provided and discussed.