Knowledge based bridge engineering - Artificial intelligence meets building information modeling

In the research project "KBE4Infra", knowledge-based engineering (KBE) techniques for capture, storage and reuse of engineering knowledge in bridge engineering are investigated. A rich data source of bridge engineering knowledge, are Bridge Management Systems (BMS). The work present addresses the question, whether it is possible to gain knowledge by the use of data in bridge management systems and how this knowledge can be used beneficially in conceptual bridge design phases. To achieve this, we propose the use of a Bayesian network. Thus, we introduce and review well known network learning and inference algorithms for Bayesian networks. Then, promising methods are matched to the needs of the proposed approach. Finally, a proof of concept will show the practicality of the chosen approach.