This paper explores the evolution of Geodesign in addressing spatial and environmental challenges from its early foundations to the recent integration of artificial intelligence (AI). AI enhances existing Geodesign methods by automating spatial data analysis, improving land use classification, refining heat island effect assessment, optimizing energy use, facilitating green infrastructure planning, and generating design scenarios. Despite the transformative potential of AI in Geodesign, challenges related to data quality, model interpretability, and ethical concerns such as privacy and bias persist. This paper highlights case studies that demonstrate the application of AI in Geodesign, offering insights into its role in understanding existing systems and designing future changes. The paper concludes by advocating for the responsible and transparent integration of AI to ensure equitable and effective Geodesign outcomes.

Tranquility is typically associated with low noise levels and remote natural areas. Various methods for preserving potentially tranquil places have been proposed, although these typically involve setting aside places with low noise levels located in remote areas. To gain the benefits of tranquility in accessible urban areas, we need to identify the characteristics of tranquil spaces. This study focuses on the landscape-based, visual aspects of the phenomena. We investigated the role of visual context using a nationwide dataset of crowdsourced photographs from Sweden. Text mining identified personal perception and accompanying photographs identified the physical features. The photographs were characterized by time period and landscape conditions using computer vision technology. We found that waterbodies consistently enhanced tranquil views, while grass, flowers, and other dense vegetation were generally not well connected. Trees were positively correlated during daylight hours but had a negative impact at night. Dynamic objects such as people and vehicles were negatively associated, potentially due to aural considerations. Their effect was less significant during hours when noise would generally be less of a factor. This study provides insights for future research and design practices aimed at promoting tranquil experiences in urban environments and demonstrates the potential for crowdsourced data to help understand the qualities of built environments as perceived by the public.

The digitization of traffic control infrastructure (TCI) is essential for modern traffic management and autonomous mobility, yet it is hindered by the fragmentation of data across commercial, governmental, and open-source platforms. This study proposes and validates a systematic framework to integrate and evaluate TCI data from these disparate sources. Focusing on speed limit, stop, and yield signs across four corridors in the Dallas-Fort Worth area, we developed a ground truth dataset for robust comparison. Our methodology involves a multi-stage process of data pre-processing and harmonization, followed by spatial consolidation using the density-based spatial clustering of applications with noise (DBSCAN) algorithm to reduce redundancy. The quality of each individual dataset and the final integrated dataset was quantitatively assessed using accuracy, completeness, and reliability metrics. Our results reveal significant reliability disparities among the sources, with the commercial dataset (Mobileye) offering the most balanced individual performance. Critically, the spatially integrated dataset achieved consistently high reliability scores (98%–100%), significantly outperforming any single source. The primary contribution of this work is a practical, replicable framework that empowers transportation agencies to systematically evaluate and integrate heterogeneous data, thereby enabling the creation of more comprehensive and trustworthy digital TCI inventories.

The rapid development of artificial intelligence in recent years, especially the rise of deep learning, provides a new path for radar target recognition. This paper first introduces the development of synthetic aperture radar and deep learning, and then analyzes the challenges and difficulties of deep neural networks. Then the self-encoder is used for target recognition. The model uses self-encoder to realize feature selection and data Manacor to classify the final object. In this paper, the advantages and disadvantages of the model are analyzed, and the improvement ideas are put forward. Finally, the original synthetic aperture radar image is processed by Median filter, mean filter and Gaussian filter. After comparison and analysis, it is concluded that the filtering effect is best when the Median filter size is 22.