A Semantic Web Approach for the Forest-based Supply Chain

In the Forest-based supply chain different stakeholders work together. In an operational supply chain data are exchanged between the individual stakeholders along with the actual products. However, media breaks may occur (Gronalt et al., 2005). Even if companies along the Forest-based supply chains may use high-sophisticated IT-solutions there is no guarantee for syntactic or semantic interoperability throughout the supply chain. One major reason therefore is that the IT architecture diverges widely across the supply chain. Hence, the stakeholders store their data in their own closed and proprietary systems, which prevents sharing of the data. Thus, collaboration between the stakeholders is neither efficient nor cost effective. However, there are new approaches that address this problem. One possible solution are semantic web approaches – in particular Resource Description Framework (RDF) and triple-stores. Although each stakeholder manages their triple store locally, triple stores offer functionalities to share and query the data including their semantics. This allows the data to be queried and visualized by all stakeholder of the supply chain (Cheng et al., 2010). This work elaborates on a prototype utilizing semantic web approaches (i.e. RDF and triple stores) to share data along the Forest-based supply chain. In particular, we focus on timber transport in the Forest-based supply chain from the forest to the saw mills. The work is based on existing standards in this area, such as Forst Holz Papier (“FHPDat”), along with previous studies relevant in that field (e.g. Schachner-Nedherer & Scholz, 2017; Scholz, 2015) which are considered during the development of a conceptual data model. The latter is designed as an ontology that is utilized for the semantic web approach thereafter. A prototype that utilizes the several triple store instances shall show the advantages of semantic web for data sharing along the supply chain. Data can be queried directly from all stakeholders using the expressive power of GeoSPARQL (Perry and Herring, 2012). The prototype is realized with test data sets from Holzcluster Styria, a regional research, development and innovation initiative in the forest industry.