SCONTO: A Supply Chain ONTOlogy That Extends and Formalizes the SCOR Model

ALICIA BOHM; HORACIO LEONE; GABRIELA HENNING

Minneapolis

Conferencia; AIChE Annual Meeting; 2011; 2011

American Institute of Chemical Engineers ( AIChE )

Nowadays, industrial supply chains (SCs), involving different actors (e.g. manufacturers, suppliers, logistics providers, etc.), are striving to boost efficiency and responsiveness. In the face of highly competitive and global markets there are big pressures to reduce lead-times, minimize logistic costs and improve customer service, by increasing agility, flexibility and responsiveness. Thus, enterprises consider supply chain management (SCM) to be a key area where improvements in the previous issues can actually be made. In this hostile environment, both an efficient management and a fluent communication among the various players are mandatory. The development of a common representation of the SC business processes, and their associated concepts, facilitates the appropriate interrelation of the various stakeholders, as well as supports correct decision making and improvement activities. The Supply Chain Operations Reference (SCOR) model captures the Supply Chain Council (SCC) consensus view of supply chain management (Supply Chain Council, 2011) from a business process perspective. It has been widely used to identify, measure, reorganize, and improve supply chain (SC) business processes. Much of its underlying content has been employed by practitioners of all over the world for many years, thus becoming a de facto standard. The SCOR model has helped managing SC related business problems through a standardized language, pre-established metrics, and common business practices, which aim at accelerating business change, improving performance, and reducing costs. Its mains goals are to support communication among supply chain partners, as well as to improve the effectiveness of SC management and its related supply chain enhancement activities. Reference models ? such as SCOR ? are abstract frameworks for identifying and defining concepts, as well as significant relationships among the entities of a domain. These frameworks, which are independent of specific technologies, implementations, or other concrete details, have several purposes. A reference model can be used as a standard, to improve the communication among people. Also, it can be employed by a community (e.g., a set of interacting companies), to define clear roles and responsibilities, and to allow the comparison of different things. Reference models can have various degrees of formalization. As these models become more formal, they turn out to be less ambiguous, more robust, contain fewer errors, and serve better to the purpose of promoting knowledge sharing, effective communication and comparison. The SCOR model has evolved and improved very much along its lifespan, until reaching the current SCOR 10.0 version, which includes online browsing facilities throughout the framework. Navigation through the SCOR model is now much easier, and linkages between business processes, metrics, practices, etc., are made more explicit. Despite these improvements, the model is so far informal; it is mostly based on textual and graphical semi-formal definitions, which may lead to some confusing interpretations. In addition, certain aspects of the model (e.g. temporal relationships among activities that are part of a workflow, the temporal perspective of metric calculations) seem to be still underspecified or omitted. Moreover, associations between the SC business processes and the performance evaluation and assessment issues are until now rather weak. This presentation will provide an overview of SCONTO, a Supply Chain ONTOlogy that takes the SCOR model as a foundation and attempts to overcome some of the limitations pointed out previously. An ontology is a formal, explicit specification of a shared conceptualization (Studer, 1998). In the last 15 years, ontologies of various types have been proposed in several domains, including the supply chain one. In the last decade, the ontology field has matured and ontologies have been recognized as key elements to reach semantic integration, becoming one of the pillars of the Semantic Web stack. Thus, the benefits of adopting an ontology not only reside in its intrinsic capacity to promote share, reuse and communicate information, but also in its ability to establish a solid foundation for interoperability and seamless data integration among distinct applications. This last capacity turns out to be of enormous importance in the context of extended supply chains, where several partners, having different information systems, need to interplay and carry out together the associated SC business processes. SCONTO is a domain ontology that has been developed during the last five years by our group. It formalizes the SC business process representation and extends the SCOR model by formally capturing the SC physical structure, its participating organizations, as well as the resources that partake in the SC business processes, their roles and flows. Besides, SCONTO includes an explicit, comprehensive, and formal representation of SC assessment related concepts (metrics and best practices). The ontology has been developed following the methodological guidelines devised by Methontology (Fernández-López et al., 1999). As the main result of its conceptualization phase, a semiformal conceptual model was built using the UML language and OCL constraints. Later, this model has been formalized and implemented in OWL. To test the ontology, different case studies were carried out, and are still being addressed. The presentation will describe the main characteristics of SCONTO´s conceptual model and its application to a particular case-study. This example will illustrate that SCONTO can unambiguously describe SC business processes and activities, their scope, associated organizational units and resources, their interrelationships, etc. For instance, it will show how the relationships between two somehow ?overlapping? processes, Source and Deliver, which are carried out by different enterprises or distinct organizational units, can be explicitly captured by the model and later analyzed for improvement. The presentation will also show how performance related concepts are formally captured within the framework of an assessment system, which explicitly models domain notions (performance dimensions and concepts, performance attributes, metrics, measures, appraisal activities, etc.) and their interrelationships. The example being discussed will also cover these aspects and will demonstrate how measures can be clearly specified (i.e. what is measured, when, where, and how something is evaluated). The talk will conclude summarizing the benefits of turning the SCOR model into an ontology by extending and formalizing it. It will also point the advantages of this approach over the ones that are based on the use of generic business process description languages, such as BPMN (OMG, 2011).