Digital building models are crucial in various applications, including planning, monitoring energy performance simulation, and structural and energy analyses. However, creating and utilizing such models remains challenging. Manual methods for generating digital building models are time-intensive and error-prone. In contrast, automated methods, such as laser scanning and point cloud technologies, provide raw environmental data but lack essential structural usage information for building elements. This research addresses these limitations by developing a systematic method to identify and classify wall elements based on their structural functionality. Such classification is pivotal for accurately interpreting the as-built condition of structures and has significant practical implications. For instance, it can assist in building renovation projects by identifying critical load-bearing elements that require preservation or reinforcement. Furthermore, these insights support redesign efforts by facilitating efficient modifications or extensions while ensuring structural integrity. To achieve these objectives, this thesis undertakes a comprehensive review of international building standards and explores the need for automated tools in the Architecture, Engineering, and Construction (AEC) industry. A wide array of parameters influencing the identification of wall structural functionality is investigated, moving beyond basic practical considerations such as wall thickness or width. This holistic approach incorporates diverse factors to enable a reliable and comprehensive classification of wall functionalities within digital building models. The proposed methodology was systematically evaluated using as-built digital building models, demonstrating promising performance. When assessed against Eurocode and International Building Code standards, the classification pipeline achieved an average accuracy of 73.91% and a recall of 94.39%. This study introduces a framework for categorizing wall elements within digital building models into distinct structural categories, including load-bearing and non load-bearing walls, based on their functional roles. The findings establish a foundation for enhancing digital modelling practices and advancing the integration of structural functionality into automated classification systems.     «

Digital building models are crucial in various applications, including planning, monitoring energy performance simulation, and structural and energy analyses. However, creating and utilizing such models remains challenging. Manual methods for generating digital building models are time-intensive and error-prone. In contrast, automated methods, such as laser scanning and point cloud technologies, provide raw environmental data but lack essential structural usage information for building elements....     »