Knowledge-graph based approach for automated selection of spare parts suitable for additive manufacturing: a railway use-case

Abstract

Spare part inventory management (SPIM) in the railway sector highly demands reliability and transparency for decentralized inventory control. Optimal SPIM should ensure the availability of needed spare parts for a service request, considering the frequency of use and criticality criterion for effective maintenance. Additive manufacturing (AM) technologies enable cost-effective production of small batch sizes often required for spare parts. However, critical component-specific information is often unstructured within engineering drawings (ED), making digital processing, and linking to existing data from enterprise resource planning (ERP) and maintenance management systems difficult. To ensure effective maintenance logistics, this paper introduces a knowledge graph (KG) that can facilitate i) interlinking multiple sources through data integration and ii) establishing a semantic data hub, thus serving as a backbone for automated assessment of component's suitability for AM. The proposed KG-based approach merges relevant (existing) ontologies, multi-structured data from ED, ERP system information, and external data sources. The approach is developed and evaluated in real-world use-cases in cooperation with the Austrian railway and public transit industry.