Search

Artificial intelligence (AI) has increasingly been integrated into various domains, significantly impacting geospatial applications. Machine learning (ML) and computer vision (CV) are critical in urban decision-making. However, urban AI implementation faces unique challenges. Academic literature on responsible AI largely focuses on general principles, with limited emphasis on the geospatial domain. This important gap in scholarly work could hinder effective AI integration in urban geospatial applications. Our study employs a multi-method approach, including a systematic academic literature review, word frequency analysis and insights from grey literature, to examine potential challenges and propose strategies for effective geospatial AI (GeoAI) integration. We identify a range of responsible practices relevant to the complexities of using AI in urban geospatial planning and its effective implementation. The review provides a comprehensive and actionable framework for responsible AI adoption in the geospatial domain, offering a roadmap for urban researchers and practitioners. It highlights ways to optimise AI benefits while minimising potential negative consequences, contributing to urban sustainability and equity.

Evaluations of plan implementation are typically conceived in terms of plan conformance(the degree to which outcomes align with planning intent) or plan performance(the degree to which decision makers apply the plan). In this research, we consider the relationship between performance–conformance by evaluating the implementation of compact activity centre policy in greater Brisbane. We examine two decades of changes to local land use planning regulations using content analysis, quantifications of permitted development intensity, and comparisons of actual land use changes with planned intent, to identify evidence of the use of metropolitan scale activity centre policy. We find that the activity centre policy performed well and was used as intended across the metropolitan area. However, despite strong performance, the policy conformed poorly in terms of achieving the expected outcomes. These findings highlight the importance of understanding plan implementation from both performance and conformance perspectives. While conformance evaluations are essential to connect planning intent to physical outcomes, understanding plan performance is also necessary to explain the results of conformance evaluations and whether conformance successes or failures were due to deficiencies with the plan itself or how the plan was used by key actors charged with its implementation.