Assembly Precedence Graph Mining Based on Similar Products

• Published in: 2022 IEEE International Conference on Industrial Technology (ICIT) pp. 1 - 7
• Main Authors: Guiza, Ouijdane, Mayr-Dorn, Christoph, Mayrhofer, Michael, Egyed, Alexander, Brandt, Heinz RiegerFrank
• Format: Conference Proceeding
• Language: English
• Published: IEEE 22.08.2022
• Subjects: assembly data inconsistencies; Assembly Line Balancing Problem; Assembly precedence graph; Assembly systems; assembly task similarity; Data mining; Geometric modeling; Industries; Limiting; Manuals; precedence constraints; precedence graph generation; Prototypes
• DOI: 10.1109/ICIT48603.2022.10002729

The Assembly line balancing Problem (ALBP) is the optimal partitioning of assembly work among stations. This optimization problem continues to be a subject of great interest to researchers. As a result, various methods were proposed relying mainly on the precedence graph as input. The assembly precedence graph is a directed acyclic graph describing the assembly technological and organizational constraints. In reality however, this data is often outdated, incomplete or altogether unavailable limiting the applicability of the available approaches to real-world assembly systems. Nonetheless, only few approaches were proposed for the generation of the precedence graph. Grounded in an industry use-case, we propose a novel approach for the mining of the assembly precedence graph relying on precedence graphs of similar products (those belonging to the same product family for example). We evaluate our approach based on real industry data of construction machine assembly including several inconsistencies. For new products (with no past feasible sequences available), our approach is able to identify 97% of the tasks independencies at a 56% precision, using as input the precedence graphs of 3 similar products.