This study evaluates the effectiveness of evacuation guidance in large transport facilities based on full-scale trials conducted at Sant Cugat Station in Barcelona. The analysis is grounded in the assumption that, under dynamically worsening fire conditions involving smoke, elevated temperature, and toxic combustion products, occupants tend to rely primarily on the “nearest-exit” heuristic, while conventional static signage has limited operational influence on route-choice behavior. Three guidance configurations were compared: static signage (T.2.1), prototype dynamic signage supported by voice messaging (ADSS; T.2.2), and an integrated Intelligent Active Dynamic Signage System (IADSS; T.3) combining a decision-support module with synchronized visual–audible communication (ADSS + DSO/PA). The methodology included a structured analysis of controlled evacuation trials conducted on the same platform, video-recording analysis, questionnaire-based assessment, and performance evaluation using T50, T95, and total evacuation time. In the static configuration, nearly all participants selected the nearest exit, confirming the dominance of intuitive route choice. The ADSS configuration improved redirection toward the target exit, but its effectiveness depended strongly on signal visibility and was reduced by inconsistent public-address messages. The strongest overall performance was observed in the integrated IADSS configuration, where the simultaneous presentation of positive and negative cues within the user’s field of view, combined with consistent voice communication, improved rerouting effectiveness and shortened the final evacuation phase, as reflected in the temporal interpretation of the evacuation curves. The findings demonstrate that effective evacuation guidance requires full information at decision nodes, continuity of guidance along the route, and strict synchronization between visual and audible communication layers.