03514ntm a2200373 i 4500 0 0 lcc 0 1 REF 20000 20000 2023-11-29 0 FKOM .K57 2023 r Thesis T000002820 2023-11-29 00:00:00 1 2023-11-29 THESIS 0 0 lcc 0 0 REF 20000 20000 2023-11-29 0 CD13485 T000002821 2023-11-29 00:00:00 1 2023-11-29 THESIS 100239 100245 MY-KuUP 20251125110824.0 t||||fr|||| 000 0 ta 231129t20232023my a|||fr|||| 000 0 eng d THE0009809 (Local) Hardback UMP eng UMP rda FKOM .K57 2023 r Thesis Kiriyadhatshini A/p Gunaratnam, author. Hazard analysis for the requirements specification of safety-critical systems using the combination of FHA and FTA techniques / Kiriyadhatshini A/p Gunaratnam Kuantan, Pahang : UMP, 2023 ©2023 xiv, 185 pages : illustrations (some color) ; 30 cm. + 1 CD-ROM rdacontent text rdacontent text rdamedia unmediated rdamedia computer rdacarrier volume rdacarrier computer disc rda text file PDF Faculty of Computing Thesis (Master of Science) -- Universiti Malaysia Pahang – 2023 Includes bibliographical references Hazard Analysis (HA) is a crucial process for identifying and mitigating risks associated with systems development. However, current HA techniques suffer from several limitations, including a lack of preliminary hazard identification and inadequate hazard documentation, which can lead to system breakdowns. Therefore, this research aims to enhance HA techniques by addressing these limitations by conducting HA in requirement specification and producing a more comprehensive hazard log. To achieve this aim, a research methodology consisting of three phases was designed. Phase 1 involved analyzing existing HA techniques and identifying gaps in hazard analysis. Phase 2 involved developing a combined hazard analysis technique that addresses these key limitations by integrating functional hazard analysis (FHA) and fault tree analysis (FTA) techniques. The proposed technique is intended for use during the requirement specification of system development to produce a comprehensive hazard log. In Phase 3, the proposed technique was evaluated through a case study of a generic patient-controlled analgesia pump model. The performance of the proposed technique was evaluated using the F1-score measure, precision, and accuracy. Four evaluation methods were used to compare the results of single FHA, single FTA, using both FHA and FTA, and combining FHA and FTA techniques. The results showed that the combined FHA and FTA technique achieved the highest performance value of 0.96 for accuracy and 0.98 for precision, recall, and F1-score measure. This concludes that though individually FHA produces a large output data while FTA is not a preliminary technique yet both of them complements each other to achieve the aim of conducting HA in requirement specification and produce a minimalized and comprehensive hazard log. Based on these findings, the combined FHA and FTA technique is recommended for implementation during the requirement specification of systems development to identify hazards and produce a comprehensive hazard log. Future directions for research could include automating the technique to identify hazards by analyzing system functions using the causal factors in terms of variables. Faculty of Computing Dissertations Universities and colleges Dissertations Theses lcc THESIS