Improvement of airworthiness certification audits of software-centric avionics systems using a cross-discipline application lifecycle management system methodology

Abstract

The use of an Application Lifecycle Management (ALM) system to promote cross-discipline data capture, tracking and traceability has shown to provide a dramatic improvement during airworthiness audits of software-centric avionics systems by reducing both the time to locate relevant information and rework associated with errors in captured data and their traceability. When configured correctly, an ALM system enables the capture of all data associated with the development of a software-centric avionics system from customer specifications through system-level requirements and design data to the lowest level of data including software source code and system and software test results. The comprehensive nature of the data capture and the integrated traceability facilitates quick and efficient retrieval of pertinent data when performing airworthiness audits and is significantly superior to more manual methods and methods that rely upon disparate tools and data sources. This paper will describe a data model that allows for the capture and tracing of cross-discipline data in an ALM system and will present data to exhibit improvements over the use of legacy systems and methods. Analysis at UTC Aerospace Systems has shown that a significant reduction in preparation and audit time, which includes activities such as creating a document and records catalog, completing a pre-audit review and summarizing traceability architectures before an audit; and locating documents and records, assessing traceability across disciplines, itemizing open problem reports and determining the current status of development and verification activities during an audit. A summary of these improvements is shown in Table 1. Organization of the data in the ALM system is a critical factor in achieving reduced search times and preventing rework associated with traceability and lost or missing records. A data model diagram is established to allow for simpler organization of the life cycle data to support product development and verification, which is used by all disciplines including Systems Engineering, Software Engineering, Systems Test Engineering, Software Test Engineering, and associated disciplines including Configuration Management, Project Management and Quality Assurance.