Semantics-aware adaptive simplification for lightweighting diverse 3D CAD models in industrial plants

Abstract

In industrial plants, an increasing need has emerged for visualization through three-dimensional (3D) models throughout all lifecycle stages. Lightweight modeling has garnered attention for improved visualization performance. The data size of an entire process plant model is enormous, and the final model consists of various forms of equipment and material model generated from diverse sources. We propose an adaptable method to simplify models with various semantics during the lifecycle of plants. The characteristics of plant equipment and materials are analyzed to define the semantics criteria, upon which the level of semantics (LOS) is defined. The functional requirements of model lightweighting in the plant industry are organized, and comprehensive lightweighting procedures for each stage satisfying those requirements are established. The proposed lightweighting procedures can be applied adaptively to the LOS of each model. We develop a prototype system to demonstrate the proposed method. We conducted simplification experiments for two assembly models with completely different LOS. Two models were simplified gradually and effectively by the proposed simplification procedures, and the data size of each model was reduced by 99 % and 83 %.