Skip to main content

CPSwarm

Periodic Reporting for period 2 - CPSwarm (CPSwarm)

Período documentado: 2018-07-01 hasta 2019-12-31

As recognized by current academic research efforts, Internet of Things (IoT) and Cyber-Physical Systems (CPS) topics refers to a single domain, which demands for established design and deployment methods. This need is pressing as CPS and Cyber-Physical Systems of Systems (CPSoS) are finding applications in several large-scale, safety-critical domains, e.g. transportation, smart cities, etc. While the increased CPS adoption has resulted in the maturation of solutions for CPS development, a single consistent science of system integration for CPS has not yet been consolidated. Therefore, CPS development remains a complex and error-prone task, often requiring a collection of separate tools. Moreover, interactions amongst CPS might lead to new behaviors and emerging properties, often with unpredictable results. Broad challenges that shall be then addressed for designing next generation CPS include: integrating complex, heterogeneous large-scale systems; interaction between human and systems; dealing with uncertainty; measuring and verifying system performance; System design.
Developing novel model-based development toolchains for CPS and swarm-enabled CPSoS would significantly stimulate the innovation capacity and reinforce competitiveness of Europe’s industry. Industries would then have the possibility to fully exploit the potential of the next generation CPS and IoT and consider new business opportunities. Bringing innovation and solving the above issues would also make further evolve the emerging scenario where people and CPS collaborate in an environment aimed at forming our sustainable future.
In this context, CPSwarm project proposes a new science of system integration and tools that pave the way towards well-established, model-based and predictive engineering design methodologies and toolchains for next generation CPS. CPSwarm tools ease development and integration of complex herds of heterogeneous CPS that collaborate based on local policies and that exhibit a collective behavior capable of solving complex, industrial-driven, real-world problems. More specifically, project results have been tested in real-world use cases in 3 different domains: swarms of drones and rovers for search and rescue applications; swarm robotics for logistic applications and autonomous driving for freight vehicles. This has allowed to demonstrate how the Workbench tools can support the user in designing such types of applications.
Overall, CPSwarm has been driven by an iterative approach tailored to better bridge the gap between requirement analysis/elicitation and development activities. The project has been subdivided in three main phases: for each phase, the partners have defined and refactored vision scenarios and use cases for the three considered application scenarios; elicited, analysed and validated requirements (also about safety and security); developed solutions to satisfy such requirements; validated these solutions and collected lessons learned that have helped to refine the requirements of the next phase. Furthermore, the business models have been defined for all the developed components.
Starting from the conceptual architecture (see the picture below), a reference architecture has been designed in several steps. The solution is based on three main components: Modelling Library, Workbench and Deployment Toolchain, which have been implemented, validated and (a set of them) released as opensource.
The Modelling Library provides reusable parts to model aspects of the CPSoS, e.g. the members, their HW characteristics and supported functionalities, swarm-enabled behaviours and the description of the relevant environment.
The Workbench includes the Modeling Tool, based on Modelio solution, which focuses on defining the swarm composition and behaviours exploiting State Machine Notation for Control Abstraction.
The Simulation and Optimization Environment offers the possibility to simulate the designed CPSoS, swarm algorithms or optimize parameters of specific swarm algorithms for CPSoS using an evolutionary approach. The Workbench also provides libraries offering automatic code generation.
The Deployment Toolchain includes two components, i.e. the Deployment Tool, supporting over-the-air deployment of generated code to target devices, and the Monitoring & Command Tool, managing the swarm mission execution and the abstraction library, based on Robotic Operating System (ROS) that allows abstracting the CPS’ functionalities.
Finally, the Launcher offers to the user a flexible access to the CPSwarm toolchain components, also promoting a proper design workflow.
Final demonstrations of the 3 scenarios have been developed.
In the dissemination area, the partners have organized telcos and physical meetings with ESG members, released main components in the opensource community and published several research papers. The communication strategy has achieved a good engagement and repercussion around CPSwarm. The international cooperation effort has led to the collaboration with the CPS Cluster’s EU projects and with the EU projects Teamplay and Brain-IoT, the participation in several international events and finally to organize the Final CPSwarm Workshop.
For exploitation, the partners have defined these actions: to make research results and CPSwarm technology specifications publicly available to allow third party to develop compatible technologies. To release CPSwarm technologies both as a commercial or opensource products. To add the know-how gained during CPSwarm to improve expertise in CPS and IoT and to create new research lines or to utilize it in internal processes.
There are many factors impeding next generation CPS system-level design, like the lack of formalized high-fidelity models for large systems, insufficient ways of measuring performance, and inadequate scientific foundations. CPSwarm mitigates these issues by providing tools and methods that pave the way towards well-established, model-based and predictive engineering design methods and toolchains for next generation CPS systems.
Considered aspects, and innovations with respect to the state of the art are manifold and encompass, for example, modeling and design methods and tools, deployment tools and code generation solutions.
The major outcomes provided by CPSwarm are:
* Drastically improve support to design of complex, autonomous CPS.
* Provide a self-contained, yet extensible, library of reusable models for describing CPS.
* Reduce in complexity and time the CPS development and deployment workflow.
* Define a complete library of swarm and evolutionary algorithms for CPS design.
* Establish reference patterns and tools for integration of CPS artefacts.
The impact of the CPSwarm design and deployment toolchain will significantly contribute to the fulfilment of the ambitious plans for Europe’s future leadership in enabling and industrial technologies. The activities have focused on the development of a new generation swarms of CPS design platform, promoting the “smart everywhere” paradigm, which will drive the convergence of CPS and IoT infrastructures. CPSwarm will also affect wider society, since complex CPSoS have a growing impact on all aspects of our lives. Indeed, beside military applications there is a clear trend to use UAV or robot swarms in civil applications (e.g. forest fire control, farming, environmental monitoring, search & rescue, surveillance, packet delivery).