Tech. Interop. International Workshop on Software Engineering for Smart yber-physical Systems (SEsPS) Versioning in yber-physical System Engineering Best-Practice and Research Agenda Richard Mordinyi and Stefan Biffl hristian-doppler Laboratory for SE Integration for Flexible Automation Systems Institute of Software Technology and Interactive Systems Vienna University of Technology http://cdl.ifs.tuwien.ac.at Tool Mec. SADA Model Mec. Tool Elec. Tool SW Analysis Workflow Model Elec. Model SW
Motivation & Background Motivation: Large-Scale Engineering s, e.g., hydro power plants, car manufacturing plants, steel mills. ooperation of different engineering disciplines. Disciplines have specific engineering tools. Manual effort required for data exchange and synchronization (high risks). Key research questions focus on: domain experts and researchers in ASE environments with a set of concepts, methods, and tools to make informed decisions on top of integrating engineering knowledge to design advanced applications for mitigating risks of defects in the engineering of flexible automation systems 2
Position of Research in PS oncept Map 3 http://cyberphysicalsystems.org
haracteristics of Automation Systems Engineering Limited engineering process analysis and improvement capabilities Engineering processes seem sequential but have loops back. Heterogeneous knowledge representations in diverse engineering models. Fragile change management in parallel multi-disciplinary engineering. Insufficient early risk management in a heterogeneous environment. 4 VDI: 'Engineering of industrial plants. Evolution and Optimizations. Part 1-4, Verein Deutscher Ingenieure e.v., VDI/VDE 3695, 2010. VDI Richtlinie 2206 Entwicklungsmethodik für mechatronische Systeme, VDI Verlag, 2004.
Industry 4.0: Engineering Knowledge at Run Time Engineering Phase Business Requirements Process Eng. Electrical Eng. AD, Pipe & Instrumentation Electrical Plan Integrate Business Requirements in Engineering -level concepts Automation Service Bus PL program Engineering ockpit Multi-Model Dashboard Software Dev. Environment ontrol Eng. Participant Software Eng. Test/Operation Phase Business Operator ERP System Planning SADA OP UA Server (augmented) -level concepts RT-Automation Service Bus Access run-time information PL program Enrich runtime information Engineering ockpit Multi-Model Dashboard Diagnosis Analysis ontrol Eng. Participant Diagnosis Expert Deploy created artifacts yber Physical System (PPS) Access engineering information yber Physical System (PPS) PL OP UA Server onfig Sales Transport Sales Transport 5 Flexibility increases system complexity Need for better integrated engineering to cope with larger solution space and with system changes at run time commissioning
Version Management of Mechatronic Objects 6 Versioning of various semantic model element levels File, Folders, Structural Elements, and detailed content levels File-Level Versioning not sufficient reflects data format syntax Detection of hanges at Model-level
Summary & Research Aspects Engineering of sps needs to cope with multiple heterogeneous engineering domains Data heterogeneity integration Methods, tools and modeling approaches of various domains Access to domain specific model data from project/process level Versioning and linking of engineering artifacts Scalability of framework in managing versions Formulation of cross-domain queries with domain-specific knowledge Model-driven engineering Modeling permitted changes of the production system during runtime Modeling corridor of allowed changes Formalization of operator s knowledge to support automation 7