<![CDATA[Purpose – This study aims to develop a web-based expert system to support initial fault identification in bus fleets, addressing the limitations of manual, experience-based diagnostics that are often subjective and time-consuming in operational environments. Design/methods/approach – The system was developed using a rule-based approach with a Decision Tree framework, where entropy and information gain were used to structure expert knowledge into an interpretable diagnostic hierarchy. The development followed the SDLC Waterfall model and incorporated 30 fault categories across six subsystems. Validation included entropy-based computation on the AC subsystem and expert-scenario testing across all subsystems (90 cases). System usability was evaluated using the System Usability Scale (SUS), and functional testing was conducted using Black Box Testing. Findings – The system achieved an accuracy of 97.78% under expert-defined diagnostic scenarios. However, this result reflects rule-consistency performance within structured scenarios and should not be interpreted as real-world diagnostic accuracy. The SUS evaluation yielded a score of 82.07, categorized as “excellent,” and all functional modules operated correctly based on Black Box Testing.Research limitations/implications – The validation is based on expert-defined scenarios rather than independently observed operational failure data, limiting generalizability. In addition, overlapping symptoms may introduce ambiguity in certain diagnostic conditions. Originality/value – This study contributes an interpretable expert system that integrates entropy-based attribute prioritization within a web-based fleet management context, providing structured diagnostic support for non-technical operational personnel.]]>
