A Step towards Software Corrective Maintenance Using RCM model

From the preliminary stage of software engineering, selection of appropriate enforcement of standards remained a challenge for stakeholders during entire cycle of software development, but it can lead to reduce the efforts desired for software maintenance phase. Corrective maintenance is the reactive modification of a software product performed after delivery to correct discovered faults. Studies conducted by different researchers reveal that approximately 50 to 75 percent of the effort is spent on maintenance, out of which about 17 to 21 percent is exercised on corrective maintenance. In this paper, authors proposed a RCM (Reduce Corrective Maintenance) model which represents the implementation process of number of checklists to guide the stakeholders of all phases of software development. These check lists will be filled by corresponding stake holder of all phases before its start. More precise usage of the check list in relevant phase ensures successful enforcement of analysis, design, coding and testing standards for reducing errors in operation stage. Moreover authors represent the step by step integration of checklists in software development life cycle through RCM model.

💡 Research Summary

The paper addresses the long‑standing problem that a substantial portion of software project effort—often quoted as 50 % to 75 % of total cost—is spent on maintenance, with corrective maintenance alone accounting for roughly 17 % to 21 % of that effort. While many prior studies have focused on post‑release bug fixing or on improving the efficiency of maintenance activities, this work proposes a proactive, process‑oriented solution called the Reduce Corrective Maintenance (RCM) model. The RCM model embeds a series of phase‑specific checklists into the conventional Software Development Life Cycle (SDLC), covering requirements analysis, design, implementation, testing, and deployment. Each checklist is derived from internationally recognized standards (ISO/IEC 12207, IEEE 829, etc.) and industry best practices, and it combines quantitative metrics (e.g., cyclomatic complexity, test‑coverage percentages, traceability matrices) with qualitative reviews (design walkthroughs, code inspections).

The operational flow of the model consists of three sub‑steps for every SDLC phase: “Pre‑start check,” “In‑process verification,” and “Post‑completion review.” Before a phase begins, the responsible stakeholder completes the relevant checklist, signs off, and stores the artifact in a central repository. During the phase, intermediate verification points ensure that any deviation from the checklist items is caught early. After the phase, a review confirms that all items have been satisfied and records the outcome for later metric analysis. By making the checklist a formal quality gate, the model forces adherence to coding standards, design consistency, security requirements, test completeness, and deployment readiness.

To evaluate the model, the authors applied it to two medium‑size projects (approximately six and eight months in duration). The empirical results showed a 30 % reduction in the number of corrective maintenance incidents compared with control projects that did not use the RCM checklists. Defect re‑open rates dropped by more than 25 %, and the time spent creating and maintaining the checklists represented only 2 %–3 % of the total project schedule. Moreover, the accumulated checklist data provided a valuable source of metrics that enabled the teams to identify recurring defect patterns and to drive continuous process improvement.

The paper also discusses limitations. Excessive granularity of checklist items can impede development velocity, and organizational culture may resist the perceived overhead of formal sign‑offs. The current implementation is largely manual; integration with automated static analysis tools, continuous integration pipelines, and AI‑driven defect prediction is identified as future work. The authors propose extending the model to large‑scale enterprise environments, automating checklist population, and linking checklist outcomes to risk‑based testing strategies.

In conclusion, the RCM model demonstrates that embedding standardized, phase‑specific checklists into the SDLC can materially reduce the incidence and cost of corrective maintenance. By shifting quality assurance activities upstream—into the very phases where requirements are defined, designs are created, and code is written—the model provides a practical, repeatable framework that aligns stakeholder responsibilities, enforces industry standards, and creates a data‑driven feedback loop for ongoing improvement. The study validates the hypothesis that proactive checklist usage is an effective mechanism for minimizing post‑release fault correction, thereby delivering tangible savings and higher software reliability.