Artifact-Centric Service Interoperation (ACSI) - overview
The ACSI EU STReP project (2010-2013), developed a data-centric service interoperation framework that is based on two novel notions: interoperation hubs and business artifacts. Interoperation hubs enable flexible, user-centric collaborations in an open network. Business artifacts provide the conceptual basis to the modeling, design, execution, and formal verification of business processes that cater across different organizations. The project, comprising a consortium of five world-class research institutes and two industrial partners, was highlighted to have gained an extraordinary scientific impact. This impact involved knowledge integration of several areas, including business process management, artifact-centric business processes, verification, data integration and ontologies, process mining, and services architectures.
Scope and Objectives
ACSI’s overall objective was to simplify the creation and maintenance of service collaboration environments. Scientifically, ACSI was aimed to develop the new notion of artifact-centric interoperation hub, and a surrounding conceptual framework of service coordination, participant views, schema evolution, verification, and process mining based on formal and empirical techniques. Technologically, ACSI was aimed to develop a substantial prototype for creating and operating interoperation hubs, that integrate the techniques devised by the scientific research stream. Two concrete use cases were used in order to demonstrate and test project results.
Achievements and Results
The ACSI project has investigated new concepts and techniques for building an IT infrastructure that is intended to support the interoperation between electronic services. Towards this end, the ACSI Artifact Abstract Model (A3M) was developed as a conceptual model of artifact-based systems in multi-actor environments that was precise enough to be fully formalized in mathematics and yet abstract enough to be instantiated in very different ways into concrete models for artifact-based systems.
On the basis of the A3M, the GSM model (Guard-Stage-Milestone) was elected and further adapted as the concrete model for the ACSI project. The decision to focus on GSM as its most prominent concrete model has led to unexpected theoretical results in the verification arena, and provided a solid basis for the OMG Case Management Model and Notation (CMMN) standard.
The main scientific achievements of the ACSI project include the development of the GSMC model checking theory and tool, the identification of run-boundedness and state-boundness as conditions for guaranteeing the decidability of verification of Data Centric Dynamic Systems (DCDS), and the development of new process-mining techniques tailored towards the GSM model. The final ACSI I-Hub prototype system gives support for the design and execution of artifact-centric applications. This involves the importing and rendering of artifact lifecycle models generated by the process mining tool - ArtifactMiner. This may then provide the baseline for a more detailed artifact-based application design, including the specification of a designated authorization-views layer that segregates application execution among multiple service providers. Finally, the design may be verified against the verification toolkit using the GSMC tool via web service invocation.
Two use cases were employed as application prototypes to verify ACSI results. The "Energy" use case was focused on the Measurement Liquidation procedure that is used in the Spanish Electric Network to calculate fair remunerations to the numerous companies participating in the energy generation, distribution and commercialization activities. The "FRIS" use case had put its focus on the initiative of the Department of Economy, Science and Innovation (EWI) that pursues to offer fast access to research information for all the stakeholders, and to increase the efficiency and effectiveness of its R&D policy. The use case focused on the applied biomedical research (TBM) funding program related artifacts at EWI’s funding agency IWT.
Several decisive advantages compared to classic BPM style modeling have been identified, among these more flexibility to changing the model in the design and prototyping phase as well as a better way to model and implement. The experience in the use cases served as a successful proof-of-concept of the GSM technology, highlighting its merits with respect to model adaptation and implementation.
The scientific productivity of ACSI, rated "excellent" by its reviewers, has led to a great impact on BPM, DB and AI research, proposing the artifact-centric approach as an opportunity for linking these different disciplines. This was indicated by numerous papers on artifact-centric approaches in the main Business Processes and Services conferences, invited talks and tutorials on ACSI results in many high-profile conferences, and by the addition of ACSI topics to the list of conference and workshop topics in these areas. Considering its scientific popularity, a designated Data- & Artifact- Centric BPM (DAB) workshop was also initiated since 2012, coinciding with the BPM conference.
The ACSI I-Hub prototype has been contributed to the FP7 project "FIspace" in the form of open source (available on sourceforge.com in the name of BizArtifact). The other ACSI technology components, including the GSMC and ArtifactMiner (part of the ProM toolkit) have been made available as open source as well.