Cyber-Physical Systems (CPS) represent one of the most prominent trends in computer technology, creating a deep integration of hardware and software components. They have evolved in conjunction with Internet of Things (IoT) and various smart approaches, in manufacturing, transportation, health care, buildings, emergency response. As the world has become increasingly inter-related - due to services, mobile devices, Cloud Computing, and wireless networks - CPS have highly proliferated, facing challenges like scalability, composability, and heterogeneity. The aim is to elaborate integrated models that drive cyber-physical systems, and specific modeling languages, for dealing with the level of complexity required in real-world settings.

Objectives

The Model-IN project supported by the DAAD Research Stay had the purpose to advance model integration for cyber-physical systems according to four perspectives:

• Research. During the Research Stay, there was the objective to perform a systematical study and a rigorous classification of integration approaches, at the level of metamodels, models and code. The analysis had to consider standards and decomposition in steps correspondent to typical software development life cycle phases. The research was expected to propose new and original methods and tools, adapted to the specific requirements of CPS. Eventually, the project would define the dimensions of integration that are most relevant for Cyber-Physical Systems.
• Applicability. The goal was to identify and integrate models for examples of CPS that are of strong interest to potential industrial partners, to respond to their real needs. Nonetheless, we wished to address applications for mastering complex cyber-physical systems, with respect to the activity lines having funding potential in binational or EU Research and Innovation programs.
• Collaboration. This is an objective not only for the duration of the scholarship, but also for medium and long-term. We believe that, working together and having fruitful discussions on cutting edge topics, we can progress much faster than individually. The current DAAD Research Stay should be a starting point for a more extended networking and for UOL collaboration with other colleagues from UPB at a subsequent stage.
• Education. The project has the objective to perform a comparative analysis, identify the current compatibilities, and propose solutions for harmonizing qualification programs from UPB with correspondent programs from UOL. The conclusions of this analysis will be used for elaborating future agreements for students and staff exchanges. As a long-term objective, the bridge established within the two months of this DAAD Research Stay will constitute a solid foundation for taking advantage of complementary expertise, to offer a more advanced support to young researchers and PhD students.

Sponsors

Model-IN is supported by DAAD (German Academic Exchange Service), within the Research Stays for University Academics and Scientists, 2018.

Common Research Agenda

Research Trends and Challenges

As the world has become increasingly inter-related – due to services, mobile devices, Cloud Computing, and wireless networks – highly distributed systems and cyberinfrastructures have proliferated, facing challenges like scalability, composability, and heterogeneity. They are based on Internet of Things , Cyber-Physical Systems and various smart approaches, applied to manufacturing, transportation, health care, buildings, emergency response, in the global context of Industry 4.0.

Due to their software intensive character, augmented by the strong interaction with the physical world, these novel systems face challenges like: security, flexibility, energy-efficiency, interactivity, and consistency. To enable new and holistic solutions, one of the highly investigated areas is Model Driven Engineering (MDE). Although “MDE 1.0” initially separated modeling from code generation, the tendencies are to overcome the borders between programming and modeling paradigms, and to get models involved in every step of the software development life cycle, as proved by approaches like: model-driven testing, model-driven systems integration, model-centric reengineering, and model-based versioning. Thus, the aim is to reach the “MDE 2.0” level, where everything is a model. Albeit this is an important research area, the limited industrial adoption of MDE shows that much more advances are necessary in technology, methods, and tools, for pervading the common industry practice, including small and medium-sized enterprises. To fulfil this goal, a contribution is certainly expected from the academic institutions, who should enable the large-scale dissemination of knowledge and the adoption of new tools; therefore, there is a need to reinforce the study of Model Driven Engineering, apply it to the development and maintenance of IoT and CPS, and to collaborate for embracing the best educational and training methods.

The complexity of these emergent systems involves the necessity to work at a high level of abstraction, therefore Model Driven Engineering represents a very promising approach; moreover, one requires multiple models, which have to be integrated both vertically - to consider all the implementation levels - and horizontally - to deal with the separation of concerns. The mixed cybernetic and physical nature, and the high non-homogeneity also lead to the necessity to define models using multiple modeling languages, and to integrate them. This introduces challenges regarding: metamodel integration, mapping, creating inter-domain relationships, standard compatibility, automate model transformations. Although the modeling of specific concerns is consecrated at the level of the state of the art, model and metamodel integration still needs innovative conceptual and software frameworks. An important contribution is also expected from Ontology Technologies, whose links with MDE have become an evidence and may be applied to assess the integration capability of systems, in respect with their application domains.

Common Research Topics

• Integration of metamodels, models and modeling languages:
  • Evaluate existing CPS / IoT languages
  • Define a framework for describing CPS / IoT languages
  • Define new methods and tools for model and metamodel integration
  • Set up a language for model-based description of CPS / IoT
  • Integrate existing tools by composing metamodels
  • Perform integration for real-world examples
  • Develop applications based on multiple modeling languages
  • Metamodel integration based on ontologies:
    • Develop applications based on both models and ontologies
    • Explore ontology-based techniques for metamodel integration and metamodel discovery
• Application of Model Driven Engineering in CPS / IoT systems development:
  • Providing tools and techniques for the realization, maintenance, and operation of CPS / IoT
  • Exploitation of new languages and tools in an industrial context
  • Cyberinfrastructures based on Internet of Things and Cyber-Physical Systems
  • Large-scale systems for environmental monitoring and hazard management, with a focus on water and air pollution
  • Smart devices in everyday life and in industrial settings

Collaboration Plan

The collaboration leans on common research interests in Software Engineering and Model Driven Engineering. There have been common research topics that were addressed separately: a model driven approach of software evolution and maintenance, metamodel integration, and modeling sensor-based environmental information systems. They set the premises for a fruitful collaboration in order to:

• Investigate more possibilities to apply model integration in industrial settings
• Analyze country-specific applicability opportunities within Small and Medium Enterprises
• Create a sustainable framework for exchanging knowledge and prepare joint research projects, facilitated by the Memorandum of Understanding between UPB and UOL
• Cooperate for teaching Software Engineering and Model Driven Engineering, and promote best students and promising young researchers in the framework of an Erasmus Plus Agreement