The advancement of predictive and preventive maintenance in automotive systems is increasingly important for improving vehicle reliability, reducing unplanned downtime, and supporting more efficient maintenance planning. Conventional maintenance approaches often rely on scheduled inspections, historical failure records, or reactive interventions after faults have occurred. These approaches may result in delayed fault detection, inefficient resource allocation, unexpected component failures, and avoidable operational disruptions. This research investigates an advanced AI-based diagnostic framework for supporting predictive and preventive maintenance in connected vehicle environments. The framework explores the integration of heterogeneous vehicle-related data sources and intelligent analytical methods to improve fault anticipation, maintenance decision support, and operational reliability. Rather than focusing only on historical failure patterns, the research aims to support more context-aware and adaptive maintenance reasoning across different vehicle operating conditions. From an academic perspective, the study contributes to the development of intelligent maintenance systems by examining how data-driven learning, semantic reasoning, and causal analytical approaches can be combined to support more reliable diagnostic and decision-making processes. From an industrial perspective, the research addresses the need for scalable, explainable, and implementation-oriented maintenance intelligence that can support automotive stakeholders, fleet operators, service providers, and technology partners. The research is positioned within a connected vehicle and Vehicle-to-Everything context, where vehicle signals, maintenance-related information, and operational indicators may be processed to support early risk identification and maintenance optimization. Certain technical components, parameterization details, and implementation mechanisms are intentionally not disclosed in this thesis introduction due to ongoing research development and potential intellectual property protection. The thesis therefore presents the conceptual, methodological, and evaluative aspects necessary for academic assessment while preserving proprietary elements that may form the basis of future industrial application or patent development.