 |
International Society of Science and Applied Technologies |
| Design, Evaluation and Implementation of Algorithms for Predicting Software Defects and Failures During the Operational Phase | ||||
| Author | Kazuhira Okumoto
|
|||
| Co-Author(s) |
|
|||
| Abstract | In a world where AI and automation are prevalent, ensuring software and hardware reliability is crucial. While defect modeling during the coding and testing phases is well-studied, there remains a gap for both theoretic and/or practical models for the operational phase. This phase, starting postdeployment (when software is in use by customers), is critical as defects can cause customer-impacting system downtime. This paper integrates approaches from software and hardware reliability modelling to propose methods for predicting operational phase defects and failures. For this phase, we propose to approximate the software failure rate as a constant and develop a digital tool for system reliability modeling, that correlate software reliability with hardware profiles for comprehensive assessments. While software and hardware reliability modelling have both been separately well-studied, this paper is a ground-breaking attempt to combine them in the context of a deployed software solution.
|
|||
| Keywords | Customer defects, software failures, operational software reliability, piecewise modelling, digital engineering | |||
| Article #: RQD2024-46 | ||||
Proceedings of 29th ISSAT International Conference on Reliability & Quality in Design |