| description abstract | The designer generates a variant product by applying several design suggestions that fulfilled a variety of customer requirements. These design suggestions rely on multiple domains of expert knowledge, which are unstructured and implicit. Moreover, these design suggestions have an impact on assembly joint information (liaison), which makes the variant design a complex problem. To effectively support the designers, this work presents a knowledge-based decision support system for assembly variant design using ontology. First, a knowledge base is built by the development of an ontology to formally represent the taxonomy, properties, and causal relationships of/among core concepts involved in the variant design. Second, a five-step sequential procedure is established to facilitate the utilization of this knowledge base for decision-making in variant design. The procedure takes the extracted liaison information from the CAD model of an existing product as the input and further used for generating a set of variant design decisions as the output through Semantic Web Rule Language (SWRL) rule-based reasoning. The inferred outputs by the process of reasoning are the design suggestions, the variant design type required for each design suggestion, and its effect on joint information. Based on the evaluation of the ontology, the precision, recall, and F-measure obtained are 79.3%, 82.1%, and 80.67%, respectively. Finally, the efficacy of the knowledge-based decision support system is evaluated using case studies from the aerospace and automotive domain. | |
