The MYRTUS Horizon Europe Project embraces the principles of the EUCloudEdgeIOT Initiative and integrates edge, fog and cloud computing platforms, leveraging a cognitive engine based on swarm intelligence and federated learning to orchestrate collaborative distributed and decentralised components.

AI-based systems differ strongly from other systems - and it is yet uncertain how this affects governing them, i.e., to which extent and how known system health management methodologies and processes need to be upgraded and partially re-invented to be fully suitable for AI-based systems. Our analysis indicates that the differences between AI-based cyber-physical systems and traditional systems impact health prognostics and management mainly due to the AI-based systems' foundation in information flows, their novel system architectures that become necessary to enable system-internal awareness to ensure that decision-making is based on accurate information, and their unique abilities, as systems that learn and thus change their behavior invalidate known health and performance indicators, often in unpredictable ways, and lifecycle management actions like updates, upgrades, or maintenance might no longer fit systems that adapted to their operational context. Based on this understanding, we propose to complement Health and Lifecycle Management methodologies for AI -based systems with a probabilistic model-based analysis of the system's information flows, giving insight into the expected timeliness and quality of information for various conditions, and thus whether the system is fit for its purpose. We see that estimating this fitness-for-purpose shows promise for various system health purposes, especially those ensuring trust in safety-critical AI applications.

Reinforcement learning (RL) algorithms interact with their environment in a trial-and-error fashion. Such interactions can be expensive, inefficient, and timely when learning on a physical system rather than in a simulation. This work develops new runtime verification techniques to predict when the learning phase has not met or will not meet qualitative and timely expectations. This paper presents three verification properties concerning the quality and timeliness of learning in RL algorithms. With each property, we propose design steps for monitoring and assessing the properties during the system's operation.

In order to bring quantum networks into the real world, we would like to determine the requirements of quantum network protocols including the underlying quantum hardware. Because detailed architecture proposals are generally too complex for mathematical analysis, it is natural to employ numerical simulation. Here we introduce NetSquid, the NETwork Simulator for QUantum Information using Discrete events, a discrete-event based platform for simulating all aspects of quantum networks and modular quantum computing systems, ranging from the physical layer and its control plane up to the application level. We study several use cases to showcase NetSquid’s power, including detailed physical layer simulations of repeater chains based on nitrogen vacancy centres in diamond as well as atomic ensembles. We also study the control plane of a quantum switch beyond its analytically known regime, and showcase NetSquid’s ability to investigate large networks by simulating entanglement distribution over a chain of up to one thousand nodes.

Composability and modularity in relation to physics are useful properties in the development of cyber-physical systems that interact with their environment. The bond-graph modeling language offers these properties. When systems structures conform to the bond-graph notation, all interfaces are defined as physical "power ports" which are guaranteed to exchange power. Having a single type of interface is a key feature when aiming for modular, composable systems. Furthermore, the facility to monitor energy flows in the system through power ports allows the definition of system-wide properties based on component properties. In this paper we present a metamodel of the bond-graph language aimed to facilitate the description and deployment of software components for cyber-physical systems. This effort provides a formalized description of standardized interfaces that enable physics-conformal interconnections. We present a use-case showing that the metamodel enables composability, reusability, extensibility, replaceability and independence of control software components.

Quantum communication brings radically new capabilities that are provably impossible to attain in any classical network. Here, we take the first step from a physics experiment to a fully fledged quantum internet system. We propose a functional allocation of a quantum network stack and construct the first physical and link layer protocols that turn ad-hoc physics experiments producing heralded entanglement between quantum processors into a well-defined and robust service. This lays the groundwork for designing and implementing scalable control and application protocols in platform-independent software. To design our protocol, we identify use cases, as well as fundamental and technological design considerations of quantum network hardware, illustrated by considering the state-of-the-art quantum processor platform available to us (Nitrogen-Vacancy (NV) centers in diamond). Using a purpose built discrete-event simulator for quantum networks, we examine the robustness and performance of our protocol using extensive simulations on a super-computing cluster. We perform a full implementation of our protocol, where we successfully validate the physical simulation model against data gathered from the NV hardware. We first observe that our protocol is robust even in a regime of exaggerated losses of classical control messages with only little impact on the performance of the system.We proceed to study the performance of our protocols for 169 distinct simulation scenarios, including tradeoffs between traditional performance metrics such as throughput and the quality of entanglement. Finally, we initiate the study of quantum network scheduling strategies to optimize protocol performance for different use cases.

Oneof themajor challengeswhendesigning software for complex systems relates to a lack of a specific and comprehensive set of rules andmethodologies. Even more so, adaptation to field conditions is difficult tomodel and implement on systems composed of a larger number of devices/components, such as distributed systems or systems of systems. On state-of-the-art technology such as wireless sensor/actuator networks and cyber-physical systems, addressing the lack of a compressive set of rules for their design and realization offers considerable benefits. If successfully realized, it can accelerate and simplify their design and implementation. The main contribution of this chapter is a clear set of rules that are specific for the design of adaptive networked embedded systems. To be more specific, we discuss design-time vs. runtime trade-offs, introduce design patterns for reconfigurable real-timemonitoring and control, propose techniques for runtime design space exploration (managing runtime reconfiguration) and a systems engineering process for runtime reconfigurable systems. We provide guidelines for all stages of the architectural process and help system and software designers in choosing wisely specific algorithms and techniques. In conclusion, this chapter introduces a set of rules (methodologies) that are specific for designing adaptive networked embedded systems. Z. Papp · C. van Leeuwen · J. de Oliveira Filho (B) TNO, Oude Waalsdorperweg 63, The Hague, The Netherlands e-mail: julio.deoliveirafilho@tno.nl C. van Leeuwen e-mail: coen.vanleeuwen@tno.nl R. del Toro Matamoros UPM, Madrid, Spain e-mail: raul.deltoro@car.upm-csic.es A. Pruteanu TU Delft, Postbus 5, Delft, The Netherlands e-mail: a.s.pruteanu@tudelft.nl P. Šůcha Czech Technical University in Prague, Technická 2, 166 27 Prague 6, Czech Republic e-mail: suchap@fel.cvut.cz © Springer Science+Business Media Singapore 2016 Z. Papp and G. Exarchakos (eds.), Runtime Reconfiguration in Networked Embedded Systems, Internet of Things, DOI 10.1007/978-981-10-0715-6_2 29

In model-based design methodologies, the system model is at the center of the development process. The model is the single source of design information, and typically is separated—logically and physically—in several views, such as requirements, specifications, design decisions, implementation guidelines, and testing protocols. When building up on such a model, the methodology prescribe—often formally—what the system model information represents, how it is manipulated and evaluated, and how it is refined from conception towards deployment. More often than not, system models and their manipulation tend to be too complex, formal, and multi-faceted to be made manually by a human designer. Such task requires appropriate tooling, that enables a correct-by-construction automation of the process. These tools then become a mandatory support to the designer. This chapter discusses how to build up a model-oriented tool chain and its most important activity flows: model editing, model visualization and transformation; model validation and evaluation; and finally, (semi-) automatic system implementation based on the system model. In addition and for each activity, emphasis goes also to the specific tool support which is necessary for the design of adaptive/reconfigurable systems.

This short paper introduces a closed-loop design optimization method for self-organizing and self-optimizing networked systems with a focus on signal processing and control. The design process starts with creating graph-based model of the system using a dedicated modelling language. The design is exported and converted to executable code in order to obtain the properties of the runtime behaviour of the system using a simulation environment. The embedding optimization loop iteratively invokes the evaluation and searches for optimal architectures and parameterization in the user defined design space. A distinguishing feature of the tool is that it allows for runtime changes in the models, i.e. it is capable of evaluating runtime reconfigurable architectures. The design space is split into two disjunct sub-spaces: one of them defines the runtime reconfigurability (the self-capabilities), the other defines the region of design time optimization. The tool is demonstrated via a real-time monitoring application.

The capability to find individuals using CCTV cameras is important for surveillance applications at large areas such as railway stations, airports and shopping centers. However, it is laborious to track and trace people over multiple cameras post incident. In this paper, we describe the live demonstration of our interactive re-identification system in a railway station. The system performs real-time track generation in multiple cameras and live re-identification and refinding of suspects which was live demonstrated in Poland. The system allows fast interactive retrieval of an individual by showing only similar candidates. An operator can find the origin or destination of a person more efficiently, especially over large time and space intervals.

The paper describes a model based approach for architecture design of runtime reconfigurable, large-scale, networked signal processing applications. A graph based modeling formalism is introduced to describe all relevant aspects of the design (functional, concurrency, hardware, communication, energy, etc.). The formalism can be used to evaluate and suggest architectural patterns for signal processing type of applications. Moreover, an architecture design evaluation tool according to the modeling formalism is described, which can place the system design under various execution scenarios (e.g. node mobility, node failure, communication link deterioration, etc.) and determine emerging system properties (such as expected lifetime, availability of services, throughput, response times, etc.). Distinguishing feature of the tool is that it allows for modeling runtime reconfiguration solutions (e.g. distributed optimization, constraint solvers) and making changes in the underlying system architecture and thus evaluate the effects of the reconfiguration. The evaluation tool can be embedded into a design optimizer, which allows an optimal trade-off between design-time and runtime configuration to be determined. The effectiveness of the approach and the tool is demonstrated via two application cases (structural integrity monitoring, greenhouse temperature distribution estimation).

Electricity generation on offshore wind parks has an increasing economic importance - the European Commission foresees that 12% of the wind energy will be produced on offshore installations by 2020, and this share is likely to increase further in the following years.

Personal navigation systems enjoy great popularity. This work presents an indoor self-localization system that resembles outdoor GPS-Positioning. An user-friendly Bluetooth-based solution has been implemented that provides the user with orientation information within a building. Bluetooth technology allows to keep a low budget and low power system. Our region-based localization method is tailored to run completely on the personal platform, and though assures user privacy. Our approach stands out with platform independence, privacy and power awareness.

A method for position determination and / or navigation of locating network nodes in the space by means of radio signal transmission between at least one locating network nodes and a plurality of reference network nodes with a known position, which are suitable for transmitting and / or receiving radio signals, characterized by the following steps: Determining the probability (Pr (d (X))) for an event (X), whether a transmitted from the first network node signal is received from the second network node or not, in the training phase, a. wherein for each point, the number of the received signals relative to the number of the emitted signals measured and from this the probability is calculated, Formation of defined regions (regions of confidence) in space with respect to at least one of the reference nodes, a. each region (Ω) comprises the points in space at which the probability (Ρ = Pr (A Ω / d)), a true statement (A Ω / d) using decision rules to meet (d), is above a certain threshold value, b. wherein the decision rules relating to the ...