MA Abschlussvortrag, Kevin Andrews, Ort: O27/545, Datum: 24.09.2014, Zeit: 10:45 Uhr

Current Business Process Management (BPM) systems are not tailored to small or medium-sized enterprises (SME) lacking expertise in BPM. This is a disadvantage for SMEs wishing to document and/or automate their business processes. There are multiple barriers that can hinder SMEs from automating their processes, such as the lack of programming skills and general understanding of typical programming language data types. Also the multitude of gateway and event types in traditional BPM systems can have a deterring effect on potential process designers. Finally, most BPM systems require deployment to on-site servers, which might not be feasable for SMEs lacking dedicated systems administrators. As a consequence this thesis and the work accompanying it define a concept for an integrated solution for simple cloud-based business process modeling and execution. This solution offers collaboration functions and rich process documentation via a modern web-interface, sketching support for rapid iterative process development, and an objectoriented (OO) data model. Created processes can be executed directly in a web-interface as part of the seamlessly integrated modeling and runtime environment. The main focus of the solution is simplicity in every aspect of the process creation and execution workflow. This thesis contributes concepts and prototypical implementations for the engine modifications that need to be applied to a typical traditional BPM engine in order to support these features. The basic concepts and their implementations can be put into the following three categories, corresponding to the main chapters of this thesis: User-customizable data objects in process models, allowing for the creation of data models that are more readable and better structured A simple programming interface allowing the use of external code in service tasks, enabling non-professional programmers to integrate their code into processes Support for the manipulation of process flow to allow for advanced simplification concepts, such as error resolution at process run-time or manually selectable XOR-gateway paths