A Framework for an artificial intelligence controlled seismic structural design system using ontologies and SysML

Abstract

The assessment and design of structure resilience under earthquakes have been a major concern for structural experts, and researchers have made improvements in terms of analytical methods, theoretical models, failure criteria, seismic reliability, and damage assessment. However, with increased knowledge sharing and collaborative work within the civil engineering industry, there is a growing interest in cross-disciplinary, multi-objective and holistic approaches to resilience design over single effort to improve structural resilience previously. Ontologies and the semantic web belonging to symbolic artificial intelligence are currently advantageous in organising multi-source information and automated data exchange, while the graphical system modelling language SysML is a state-of-the-art system tool for designing interdisciplinary tasks. However, civil engineering professionals have little exposure to knowledge engineering and systems engineering and lack the foundation to build joint work. This study therefore proposes a framework based on ontologies and SysML for a fully machine-controlled structural seismic system. Through ontologies, SysML graphical workflows, AI code and artifacts, we can construct command streaming that understands semantics, actively searches and bridges different disciplinary contexts, giving computers a simulated sense of autonomy to understand and perform cross-domain tasks without human intervention. The proposed architecture, when applied to multidisciplinary involvement in structural resilience design, allows the AI to control and integrate isolated seismic components