Emergency smoke evacuation systems in buildings in case of fire: From natural and mechanical approaches to artificial intelligence-driven strategies

Smoke movement and accumulation during building fires remain a dominant life-safety threat, as tragically demonstrated by recent large-scale incidents such as the 2025 Kartalkaya hotel fire. Beyond obstructing visibility, smoke rapidly degrades breathable air quality, directly constraining evacuation and firefighting effectiveness. To mitigate these risks, smoke management and smoke control systems implemented through natural, mechanical and hybrid ventilation configurations are widely employed to regulate smoke stratification and preserve tenable conditions. This review provides a comprehensive and critical synthesis of smoke ventilation systems, examining their classification, design principles and operational performance across diverse building typologies. International regulatory frameworks, including EN 12101, NFPA 92 and ISO 21927, are analysed alongside performance-based design methodologies, advanced modelling tools and emerging control strategies such as artificial intelligence, sensor-driven automation and digital twins. Particular attention is given to the interaction between ventilation performance, evacuation dynamics and system resilience under variable environmental and operational conditions. Unlike conventional reviews that primarily catalogue technologies or standards, this study reframes smoke ventilation as a governing life-safety system that directly defines Available Safe Egress Time through its coupling with pressure zoning, smoke stratification stability and adaptive control logic. By positioning smoke ventilation at the core of performance-based fire safety design, the review clarifies application boundaries, identifies scenario-dependent optimisation challenges and highlights pathways for integrating intelligent control within regulatory-compliant frameworks. The findings support a shift from prescriptive system selection towards context-specific performance optimisation, providing actionable insights for researchers, engineers and policymakers engaged in the development of resilient and adaptive smoke ventilation strategies.