<p>In recent years, cities have experienced rapid urban growth, which negatively impacts the accessibility of their inhabitants. Smart cities and intelligent transportation systems (ITS) aim to optimize urban service access by providing accurate public transportation information. Traditional spatial approaches for evaluating accessibility are primarily static and fail to incorporate temporal and spatial variations in travel conditions, which can lead to inaccurate estimates of travel times and accessibility. The authors of this article propose a new dynamic spatial approach based on Geographic Information System (GIS) techniques and artificial intelligence (AI), for predicting travel times according to spatiotemporal fluctuations. The study is applied to the new city of Ali Mendjeli in Constantine, Algeria, which faces significant challenges in terms of congestion and excessive travel times. The findings revealed a notable enhancement in the accuracy of accessibility assessments, with the dynamic model continuously adjusting predictions based on current traffic conditions and temporal patterns, encompassing weekdays and weekends. Notably, during rush hour, the predictions of the dynamic model exhibited a close alignment with the actual travel times. In contrast to a static approach, this provides a more comprehensive and reliable representation of urban travel patterns. The broader implications of this study extend to other cities experiencing rapid urbanization, traffic congestion, or exhibiting complex travel patterns. The study contributes to the optimization of public transportation planning through a data-driven, responsive framework. Furthermore, the proposed approach promotes the usage of public transportation, thus reducing the environmental impact of mobility. This dual focus on efficiency and sustainability corresponds with broader objectives of implementing smart city initiatives.</p>