Autonomy has emerged as one of the most foundational and contested concepts underpinning modern intelligent systems, spanning philosophy, control engineering, robotics, artificial intelligence, transportation, and cyber-physical infrastructures. Despite decades of research, autonomy remains theoretically fragmented, often treated either as a purely technical capability or as a metaphysical attribute linked to free will and agency. This article develops a comprehensive, interdisciplinary, and theoretically grounded examination of autonomy as a multidimensional construct. Drawing strictly on the provided body of literature, the paper integrates philosophical interpretations of autonomy and free will with engineering-centric perspectives from control theory, robotics, and intelligent systems. The study systematically elaborates how autonomy evolved from classical control paradigms to intelligent and adaptive architectures, and how these developments reshape human–machine interaction, ethical accountability, and system governance. Particular attention is given to levels of autonomy, agent-based models, emotion and embodiment in autonomous behavior, and the transition from rule-based automation to learning-enabled self-regulating systems. In addition, the article extends the discussion to contemporary cyber-physical and vehicular systems, including autonomous vehicles, Internet of Vehicles, and secure networked autonomy, situating these within broader socio-technical and infrastructural contexts. Methodologically, the research adopts a qualitative, theory-synthesis approach, performing an extensive conceptual integration and comparative analysis of foundational and applied works. The results reveal that autonomy cannot be reduced to independence or automation alone but must be understood as an emergent property arising from perception, decision-making, learning, control, interaction, and contextual embeddedness. The discussion critically examines limitations in current autonomy frameworks, including oversimplified taxonomies, insufficient consideration of emotion and ethics, and gaps between philosophical autonomy and engineered autonomy. The article concludes by proposing a unified conceptual lens that positions autonomy as a continuum rather than a binary property, emphasizing its dynamic, relational, and system-level nature. This work contributes a deeply elaborated theoretical foundation intended to support future research, system design, and policy-making in intelligent autonomous systems.