Using STPA and CAST to design for serviceability and diagnostics

Abstract

OEM industries are facing increased challenges providing proactive and reactive equipment support. Increased product complexity and the fast rate of technology change make problems difficult to understand, prevent, and resolve. The cost of machine unavailability is extreme, and reliability-based design methods ignore service time as a key contributor to machine unavailability. Serviceability and diagnostics are an important control to minimize customer losses when problems do occur. Methods are needed that identify serviceability needs early in the product development process while managing product complexity. STAMP (System-Theoretic Accident Model and Processes) is an accident causality model developed as a new engineering approach to system safety. While it was originally created for safety, its foundation in systems theory lends itself to other emergent properties, like serviceability. This research demonstrates that STAMP techniques can be applied to address existing serviceability issues and to guide service-friendly system design in early, conceptual design phases. Two case studies, drawn from industry, are explored to verify the effectiveness of applying STAMP to serviceability. Both case studies successfully generated hardware, software, and operator interface design requirements. They also produced recommendations for the product development and support processes. By using STAMP techniques to understand system interactions and strengthen service control structures, OEMs can address many of the challenges they are currently facing providing serviceability and support.