NTAB0: Design Priors for AI-Augmented Generative Design of Network Tied-Arch-Bridges

Projects in the Architecture, Engineering and Construction (AEC) industry inherit a great complexity due to a tremendous amount of design parameters, multiple objectives, and many involved stakeholders. Especially in the conceptual design stage of bridges, an in-detail analysis of many performance attributes for each design alternative is time-consuming and infeasible under the current approaches. In the industry today, therefore the initial design solution predominantly depends on the expertise of the involved team. In contrast to the status quo, this paper introduces the novel concept of bridge design prior models to predict the layout and structural properties of bridges as the (near-optimal) starting point for Generative Design. The concept of design prior models for bridges is demonstrated on network tied-arch bridges (NTAB). NTAB0 is calibrated upon a curated database consisting of existing real-world NTABs and captures numeric, semantic, and topological relations between bridge properties such as materials, cross-sections or bracing systems. First, a clustering analysis is performed by applying the k-Prototype and DBSCAN algorithms. In the second step, a predictive model is trained using a gradient-boosted decision tree algorithm. A subsequent study evaluates the suitability of the algorithms to serve as sensible design priors. We found that the AI prior model NTAB0 is able to suggest meaningful design parameters, assisting the designing team with an informed initial bridge design for further design space exploration and optimisation. It enables designers to make more informed decisions towards optimised bridge structures at an early …