Cyber-Physical Systems (CPSes) form the core of Industrial Control Systems (ICS) and critical infrastructures. These systems use computers to control and monitor physical processes in many critical industries including aviation, industrial automation, transportation, communications, waste treatment, and power systems. Increasingly, these systems are connected with corporate networks and the Internet, making them susceptible to risks similar to traditional computing systems experiencing cyber-attacks on a conventional IT network. Furthermore, recent attacks like the Stuxnet worm have demonstrated the weaknesses of CPS security, which has gained much attention in the research community to develop more effective security mechanisms. While this remains an important topic of research, often CPS security is not given much attention in undergraduate programs. There can be a significant disconnect between control system engineers with CPS engineering skills and network engineers with an IT background. This thesis describes hands-on courseware to help students bridge this gap. This courseware incorporates cyber-physical security concepts into effective learning modules that highlight real-world technical issues. A modular learning approach helps students understand CPS architectures and their vulnerabilities to cyber-attacks via experiential learning, and acquire practical skills through actively participating in the hands-on exercises. The ultimate goal of these lab modules is to show how an adversary would break into a conventional CPS system by exploiting various network protocols and security measures implemented in the system. A mock testbed environment is created using commercial-off-the-shelf hardware to address the unique aspects of a CPS, and serve as a cybersecurity trainer for students from control system or IT backgrounds. The modular nature of this courseware, which uses an economical and easily replicable hardware testbed, make this experience uniquely available as an adjunct to a conventional embedded system, control system design, or cybersecurity courses. To assess the impact of this courseware, an evaluation survey is developed to measure the understanding of the unique aspects of CPS security addressed. These modules leverage the existing academic subjects, help students understand the sequence of steps taken by adversaries, and serve to bridge theory and practice. / Master of Science

The field of robotics has witnessed significant advancements in recent years, with robotic arms playing a pivotal role in various industrial and research applications. In large-scale manufacturing, manual labour has been replaced with robots due to their efficiency in time and cost. However, in order to replace human labour, the robots need to collaborate in a way that humans do. This master's thesis, conducted at the Cyber-physical Systems Lab (CPS-Lab) at Uppsala University, delves into the intricacies of cooperative interactions between two homogenous robotic arms powered by machine learning algorithms, aiming to explore their collective capabilities. The project will focus on implementing a multi-agent cart-pole experiment that will challenge the two robotic arms' cooperative abilities. First, the problem is simulated, and afterwards implemented in real life. The experiment will be evaluated by the performance of various tested machine learning algorithms. In the end, The simulation yielded poor results due to the complexity of the problem and the lack of proper hyperparameter tuning. The real life experiment failed instantly, caused by the robotic arms not being designed for this application, a large simulation gap, and latency in the controller design. Overall, the results show that the experiment was challenging for the robotic arms, but that it might be possible under different circumstances.

Remote control of automation systems is consistently undeniably as a crucial aspect of their development, as it eliminates the need to travel unnecessary distances to operate them. Therefore, a framework is proposed not only for controlling an industrial robotic system but also for monitoring its behaviour and environment to ensure efficient and secure control over it. This project is carried out within the field of robotics, although its application can extend to other domains such as automotive, among others.  In the following project, a system based on industry 5.0 and Cyber Physical Systems is developed and implemented capable of storing and recovering the data collected from a robotic station while allowing its control through a User Interface. Giving the operator the opportunity to control an industrial assembly process remotely in a reliable and safe way.

Doctor of Philosophy / Computing and Information Sciences / Scott A. DeLoach / As sensor prices drop and computing devices continue to become more compact and powerful, computing capabilities are being embedded throughout our physical environment. Connecting these devices in cyber-physical systems (CPS) enables applications with significant societal impact and economic benefit. However, engineering CPS poses modeling, architecture, and engineering challenges and, to fully realize the desired benefits, many outstanding challenges must be addressed. For the cyber parts of CPS, two decades of work in the design of autonomous agents and multiagent systems (MAS) offers design principles for distributed intelligent systems and formalizations for agent-oriented software engineering (AOSE). MAS foundations offer a natural fit for enabling distributed interacting devices. In some cases, complex control structures such as holarchies can be advantageous. These can motivate complex organizational strategies when implementing such systems with a MAS, and some designs may require agents to act in multiple groups simultaneously. Such agents must be able to manage their multiple associations and assignments in a consistent and unambiguous way. This thesis shows how designing agents as systems of intelligent subagents offers a reusable and practical approach to designing complex systems. It presents a set of flexible, reusable components developed for OBAA++, an organization-based architecture for single-group MAS, and shows how these components were used to develop the Adaptive Architecture for Systems of Intelligent Systems (AASIS) to enable multigroup agents suitable for complex, multigroup MAS. This work illustrates the reusability and flexibility of the approach by using AASIS to simulate a CPS for an intelligent power distribution system (IPDS) operating two multigroup MAS concurrently: one providing continuous voltage control and a second conducting discrete power auctions near sources of distributed generation.

Doctor of Philosophy / Department of Electrical and Computer Engineering / Don Gruenbacher / Caterina Scoglio / Cyber physical systems emerge when physical systems are integrated with communication networks. In particular, communication networks facilitate dissemination of data among components of physical systems to meet key requirements, such as efficiency and reliability, in achieving an objective. In this dissertation, we consider one of the most important cyber physical systems: the smart grid. The North American Electric Reliability Corporation (NERC) envisions a smart grid that aggressively explores advance communication network solutions to facilitate real-time monitoring and dynamic control of the bulk electric power system. At the distribution level, the smart grid integrates renewable generation and energy storage mechanisms to improve reliability of the grid. Furthermore, dynamic pricing and demand management provide customers an avenue to interact with the power system to determine electricity usage that satisfies their lifestyle. At the transmission level, efficient communication and a highly automated architecture provide visibility in the power system; hence, faults are mitigated faster than they can propagate. However, higher levels of reliability and efficiency rely on the supporting physical communication infrastructure and the network technologies employed. Conventionally, the topology of the communication network tends to be identical to that of the power network. In this dissertation, however, we employ a Demand Response (DR) application to illustrate that a topology that may be ideal for the power network may not necessarily be ideal for the communication network. To develop this illustration, we realize that communication network issues, such as congestion, are addressed by protocols, middle-ware, and software mechanisms. Additionally, a network whose physical topology is designed to avoid congestion realizes an even higher level of performance. For this reason, characterizing the communication infrastructure of smart grids provides mechanisms to improve performance while minimizing cost. Most recently, algebraic connectivity has been used in the ongoing research effort characterizing the robustness of networks to failures and attacks. Therefore, we first derive analytical methods for increasing algebraic connectivity and validate these methods numerically. Secondly, we investigate impact on the topology and traffic characteristics as algebraic connectivity is increased. Finally, we construct a DR application to demonstrate how concepts from graph theory can dramatically improve the performance of a communication network. With a hybrid simulation of both power and communication network, we illustrate that a topology which may be ideal for the power network may not necessarily be ideal for the communication network. To date, utility companies are embracing network technologies such as Multiprotocol Label Switching (MPLS) because of the available support for legacy devices, traffic engineering, and virtual private networks (VPNs) which are essential to the functioning of the smart grid. Furthermore, this particular network technology meets the requirement of non-routability as stipulated by NERC, but these benefits are costly for the infrastructure that supports the full MPLS specification. More importantly, with MPLS routing and other switching technologies, innovation is restricted to the features provided by the equipment. In particular, no practical method exists for utility consultants or researchers to test new ideas, such as alternatives to IP or MPLS, on a realistic scale in order to obtain the experience and confidence necessary for real-world deployments. As a result, novel ideas remain untested. On the contrary, OpenFlow, which has gained support from network providers such as Microsoft and Google and equipment vendors such as NEC and Cisco, provides the programmability and flexibility necessary to enable innovation in next-generation communication architectures for the smart grid. This level of flexibility allows OpenFlow to provide all features of MPLS and allows OpenFlow devices to co-exist with existing MPLS devices. Therefore, in this dissertation we explore a low-cost OpenFlow Software Defined Networking solution and compare its performance to that of MPLS. In summary, we develop methods for designing robust networks and evaluate software defined networking for communication and control in cyber physical systems where the smart grid is the system under consideration.

La gestion des suites chirurgicales joue un rôle central pour permettre aux hôpitaux d’offrir l’accès aux soins à des coûts raisonnables. L'informatisation et l'automatisation sont des évolutions conventionnelles pour améliorer l’efficacité. Toutefois, un soutien inadapté ne peut améliorer l'activité de gestion et peut nuire à son action. Notre hypothèse est que des fonctionnalités interactives, utilisables, flexibles et adaptée aux spécificités des activités locales peuvent créer un environnement de travail dans lequel le personnel médical est capable de réagir à des événements inattendus et de s’approprier la technologie. Nos contributions comprennent en une analyse de l'activité de l'équipe chirurgicale, basée sur des entretiens, observations, une revue de la littérature et une analogie avec l'aviation civile. Nous avons participé à la construction d'un modèle mathématique du flux chirurgical et d'une visualisation de ce modèle. Nous avons identifié les exigences et principes de conception nécessaires au développement, à l'intégration et à l'appropriation d'un outil pour soutenir la gestion du flux chirurgical. Nous avons conçu des interactions multi-utilisateurs sur une grande surface et développé un prototype de tableau blanc électronique, OnBoard, qui démontre l'intégration des spécifications et des défis techniques. OnBoard appartient à un système cyber-physique comprenant des capteurs dans les salles d'opération. Enfin, nous avons déployé et évalué OnBoard dans une suite chirurgicale. L'expérience de OnBoard suggère que la conception des interactions est primordiale pour offrir un environnement collaboratif efficace au personnel médical. / Surgical suite management plays a key role in the endeavor of hospitals: patients’ health at sustainable cost. Computerization and automation of processes are conventional solutions to support resource management and efficiency. However, unsuitable support might not improve the management activity, and can even be detrimental to it. Our hypothesis is that usable and flexible interactivity tuned to local particularities can create a working environment in which the medical staff can cope with unexpected surgery events and appropriate the technology. Our contributions comprise an analysis of the activity of the surgical team, based on interviews, observations, review of the literature and an analogy with civil aviation. We participated in the construction of a mathematical model of the surgical workflow and a visualization of the mathematical model. We conducted an experimentation to identify bottlenecks of workflow inefficiencies and delays. We identified scenarios, requirements and design principles necessary to the development, integration and acceptation of a tool to support surgical workflow activities. We designed multi-users interactions on a large surface and made a prototype of electronic whiteboard, OnBoard, for the surgical suite which demonstrates the integration of the specifications and technical challenges. OnBoard belongs to a larger cyber physical system including activity sensors in every operating room of the surgical suite. Finally, we deployed the prototype in a surgical suite and evaluated it. The OnBoard experience suggests that the design of interactions is paramount to provide the medical staff an efficient collaborative environment.

<div>This dissertation addresses two significant issues in the effective application of machine learning algorithms and models for the physical and engineering sciences. The first is the broad challenge of automated modeling of data across different processes in a physical system. The second is the dilemma of obtaining insightful interpretations on the relationships between the inputs and outcome of a system as inferred from complex, black box machine learning models.</div><div><br></div><div><b>Automated Geometric Shape Deviation Modeling for Additive Manufacturing Systems</b></div><div><b><br></b></div><div>Additive manufacturing systems possess an intrinsic capability for one-of-a-kind manufacturing of a vast variety of shapes across a wide spectrum of processes. One major issue in AM systems is geometric accuracy control for the inevitable shape deviations that arise in AM processes. Current effective approaches for shape deviation control in AM involve the specification of statistical or machine learning deviation models for additively manufactured products. However, this task is challenging due to the constraints on the number of test shapes that can be manufactured in practice, and limitations on user efforts that can be devoted for learning deviation models across different shape classes and processes in an AM system. We develop an automated, Bayesian neural network methodology for comprehensive shape deviation modeling in an AM system. A fundamental innovation in this machine learning method is our new and connectable neural network structures that facilitate the transfer of prior knowledge and models on deviations across different shape classes and AM processes. Several case studies on in-plane and out-of-plane deviations, regular and free-form shapes, and different settings of lurking variables serve to validate the power and broad scope of our methodology, and its potential to advance high-quality manufacturing in an AM system.</div><div><br></div><div><b>Interpretable Machine Learning</b></div><div><b><br></b></div><div>Machine learning algorithms and models constitute the dominant set of predictive methods for a wide range of complex, real-world processes. However, interpreting what such methods effectively infer from data is difficult in general. This is because their typical black box natures possess a limited ability to directly yield insights on the underlying relationships between inputs and the outcome for a process. We develop methodologies based on new predictive comparison estimands that effectively enable one to ``mine’’ machine learning models, in the sense of (a) interpreting their inferred associations between inputs and/or functional forms of inputs with the outcome, (b) identifying the inputs that they effectively consider relevant, and (c) interpreting the inferred conditional and two-way associations of the inputs with the outcome. We establish Fisher consistent estimators, and their corresponding standard errors, for our new estimands under a condition on the inputs' distributions. The significance of our predictive comparison methodology is demonstrated with a wide range of simulation and case studies that involve Bayesian additive regression trees, neural networks, and support vector machines. Our extended study of interpretable machine learning for AM systems demonstrates how our method can contribute to smarter advanced manufacturing systems, especially as current machine learning methods for AM are lacking in their ability to yield meaningful engineering knowledge on AM processes. <br></div>

A Indústria 4.0, ou quarta revolução industrial, é o atual cenário industrial que estabelece um novo paradigma para os sistemas de produção. A indústria 4.0 é compreendida como a implementação da fábrica inteligente que opera de forma mais autônoma e com menor intervenção humana, cujo propósito é prover serviços e produtos inteligentes que atendam às necessidades individuais dos consumidores. A Indústria 4.0 está amparada nos sistemas ciber-físicos (CPS) e na Internet das Coisas (IoT). Neste cenário máquinas e produtos se comunicam entre si visando automatizar os processos industriais por meio de informações individuais obtidas em tempo real durante os processos de manufatura. No entanto, a Indústria 4.0 e as pesquisas em torno desse assunto ainda são muito recentes e requerem mais investigações no que diz respeito às arquiteturas que suportem a sua implementação, entre elas a comunicação entre produtos e máquinas. Neste quesito, recentemente foi proposto o modelo de arquitetura de referência para a Indústria 4.0 (RAMI 4.0) com o objetivo de nortear a implementação deste tipo de sistema. Contudo, o RAMI 4.0 ainda requer esforços no campo da pesquisa sob diferentes aspectos, entre eles a integração vertical de recursos do sistema de produção. Neste sentido, este trabalho objetiva apresentar uma arquitetura para a descoberta de equipamentos para processar operações conforme as necessidades dos produtos. A arquitetura foi projetada em camadas baseadas no RAMI 4.0 para prover componentes que permitam a comunicação entre equipamentos e produtos, e um mecanismo similar ao sistema de nomes de domínios (DNS - Domain Name System) para realizar a descoberta de equipamentos para processar uma determinada operação. Nessa arquitetura as informações dos equipamentos são armazenadas em uma estrutura organizada hierarquicamente para auxiliar o serviço de descoberta, e os produtos possuem informações das operações necessárias para o processo de manufatura. Para garantir a eficácia do funcionamento dos componentes e suas interações, é necessário a verificação e validação por meio de métodos formais. Neste trabalho a verificação e validação é realizada por meio da técnica PFS (Production Flow Schema)/RdP (Rede de Petri). Por fim, a arquitetura é aplicada em um sistema de produção modular para demonstrar a sistemática de implementação e a sua efetividade. / The Industry 4.0, also known as fourth industrial revolution, is the current industrial scenario that sets a new paradigm for production systems. The Industry 4.0 can be understood as the implementation of the smart factory that operates more autonomously and with less human intervention. The purposes of it is to provide smart products and services that meet the consumer individual needs. The Industry 4.0 is supported by cyber-physical systems (CPS) and Internet of Things (IoT). In this scenario machines and products communicate with each other to automate industrial processes through individual information that are obtained in real time during manufacturing processes. However, the researches around this issue are still very recent and require further investigations with regard of to the architectures that support its implementation, including communication between products and equipment. Taking into account this problem, a Reference Architectural Model for Industry 4.0 (RAMI 4.0) was recently proposed with the purpose to guide the implementation of this system type. However, the RAMI 4.0 still requires efforts in different aspects, including the vertical integration of resources of the production systems. In this sense, this work aims to present an architecture for the discovery of equipment to process operations according to the product needs. The architecture was designed based on layers of the RAMI 4.0 to provide components that allow communication between equipment and products and a Web Service that offer a mechanism similar to the Domain Name System (DNS) to locate equipment to process a required operation. In this architecture the capable operations supported by the equipment are stored in a structure organized hierarchically to aid the discovery service, and the products have information of the operation required for the manufacturing process. In order to guarantee the effectiveness of the component functionalities and their interactions it is necessary to verify and validate them by formal methods. In this work the Production Flow Schema (PFS)/Petri Net (PN) technique is used to develop the conceptual and functional modeling of the architecture. Finally the architecture is applied in a modular production system to demonstrate its implementation systematics and its effectiveness.

Les systèmes cyber-physiques mélangent des comportements physiques continus, tel la vitesse d'un véhicule, et des comportement discrets, tel que le régulateur de vitesse d'un véhicule. Ils sont désormais omniprésents dans notre société. Un grand nombre de ces systèmes sont dits critiques, i.e. une mauvaise conception entraînant un comportement non prévu, un bug, peut mettre en danger des êtres humains. Il est nécessaire de développer des méthodes pour garantir le bon fonctionnement de tels systèmes. Les méthodes formelles regroupent des procédés mathématiques pour garantir qu'un système se comporte comme attendu, par exemple que le régulateur de vitesse n'autorise pas de dépasser la vitesse maximale autorisée. De récents travaux ont permis des progrès significatifs dans ce domaine, mais l'approche adoptée est encore monolithique, i.e. que le système est modélisé d'un seul tenant et est ensuite soumis à la preuve. Notre problématique est comment modéliser efficacement des systèmes cyber-physiques dont la complexité réside dans une répétition de morceaux élémentaires. Et une fois que l'on a obtenu une modélisation, comment garantir le bon fonctionnement de tels systèmes. Notre approche consiste à modéliser le système de manière compositionnelle. Plutôt que de vouloir le modéliser d'un seul tenant, il faut le faire morceaux par morceaux, appelés composants. Chaque composant correspond à un sous-système du système final qu'il est simple de modéliser. On obtient le système complet en assemblant les composants ensembles. Ainsi une usine de traitement des eaux est obtenue en assemblant différentes cuves. L'intérêt de cette méthode est qu'elle correspond à l'approche des ingénieurs dans l'industrie : considérer des éléments séparés que l'on compose ensuite. Mais cette approche seule ne résout pas le problème de la preuve de bon fonctionnement du système. Il faut aussi rendre la preuve compositionnelle. Pour cela, on associe à chaque composant des propriétés sur ses entrées et sortie, et on prouve qu'elles sont respectées. Cette preuve peut être effectué par un expert, mais aussi par un ordinateur si les composants sont de tailles raisonnables. Il faut ensuite nous assurer que lors de l'assemblage des composants, les propriétés continuent à être respectées. Ainsi, la charge de la preuve est reportée sur les composants élémentaires, l'assurance du respect des propriétés désirées est conservée lors des étapes de composition. On peut alors obtenir une preuve du bon fonctionnement de systèmes industriels avec un coût de preuve réduit. Notre contribution majeure est de proposer une telle approche compositionnelle à la fois pour modéliser des systèmes cyber-physiques, mais aussi pour prouver qu'ils respectent les propriétés voulues. Ainsi, à chaque étape de la conception, on s'assure que les propriétés sont conservées, si possible à l'aide d'un ordinateur. Le système résultant est correct par construction. De ce résultat, nous avons proposé plusieurs outils pour aider à la conception de systèmes cyber-physiques de manière modulaire. On peut raisonner sur les propriétés temporelles de tels systèmes, par exemple est-ce que le temps de réaction d'un contrôleur est suffisamment court pour garantir le bon fonctionnement. On peut aussi raisonner sur des systèmes où un mode nominal cohabite avec un mode d'urgence. / Cyber-physical systems mix continuous physical behaviors, e.g. the velocity of a vehicle, and discrete behaviors, e.g. the cruise-controller of the vehicle. They are pervasive in our society. Numerous of such systems are safety-critical, i.e. a design error which leads to an unexpected behavior can harm humans. It is mandatory to develop methods to ensure the correct functioning of such systems. Formal methods is a set of mathematical methods that are used to guarantee that a system behaves as expected, e.g. that the cruise-controller does not allow the vehicle to exceed the speed limit. Recent works have allowed significant progress in the domain of the verification of cyber-physical systems, but the approach is still monolithic. The system under consideration is modeled in one block. Our problematic is how to efficiently model cyber-physical systems where the complexity lies in a repetition of elementary blocks. And once this modeling done, how guaranteeing the correct functioning of such systems. Our approach is to model the system in a compositional manner. Rather than modeling it in one block, we model it pieces by pieces, called components. Each component correspond to a subsystem of the final system and are easier to model due to their reasonable size. We obtain the complete system by assembling the different components. A water-plant will thus be obtained by the composition of several water-tanks. The main advantage of this method is that it corresponds to the work-flow in the industry : consider each elements separately and compose them later. But this approach does not solve the problem of the proof of correct functioning of the system. We have to make the proof compositional too. To achieve it, we associate to each component properties on its inputs and outputs, then prove that they are satisfied. This step can be done by a domain expert, but also by a computer program if the component is of a reasonable size. We have then to ensure that the properties are preserved through the composition. Thus, the proof effort is reported to elementary components. It is possible to obtain a proof of the correct functioning of industrial systems with a reduced proof effort. Our main contribution is the development of such approach in Differential Dynamic Logic. We are able to modularly model cyber-physical systems, but also prove their correct functioning. Then, at each stage of the design, we can verify that the desired properties are still guaranteed. The resulting system is correct-by-construction. From this result, we have developed several tools to help for the modular reasoning on cyber-physical systems. We have proposed a methodology to reason on temporal properties, e.g. if the execution period of a controller is small enough to effectively regulate the continuous behavior. We have also showed how we can reason on functioning modes in our framework.

Cyber-physical systems achieve sophisticated system behavior exploring the tight interconnection of physical coupling present in classical engineering systems and information technology based coupling. A particular challenging case are systems where these cyber-physical systems are formed ad hoc according to the specific local topology, the available networking capabilities, and the goals and constraints of the subsystems captured by the information processing part. In this paper we present a formalism that permits to model the sketched class of cyber-physical systems. The ad hoc formation of tightly coupled subsystems of arbitrary size are specified using a UML-based graph transformation system approach. Differential equations are employed to define the resulting tightly coupled behavior. Together, both form hybrid graph transformation systems where the graph transformation rules define the discrete steps where the topology or modes may change, while the differential equations capture the continuous behavior in between such discrete changes. In addition, we demonstrate that automated analysis techniques known for timed graph transformation systems for inductive invariants can be extended to also cover the hybrid case for an expressive case of hybrid models where the formed tightly coupled subsystems are restricted to smaller local networks. / Cyber-physical Systeme erzielen ihr ausgefeiltes Systemverhalten durch die enge Verschränkung von physikalischer Kopplung, wie sie in Systemen der klassichen Igenieurs-Disziplinen vorkommt, und der Kopplung durch Informationstechnologie. Eine besondere Herausforderung stellen in diesem Zusammenhang Systeme dar, die durch die spontane Vernetzung einzelner Cyber-Physical-Systeme entsprechend der lokalen, topologischen Gegebenheiten, verfügbarer Netzwerkfähigkeiten und der Anforderungen und Beschränkungen der Teilsysteme, die durch den informationsverabeitenden Teil vorgegeben sind, entstehen. In diesem Bericht stellen wir einen Formalismus vor, der die Modellierung der eingangs skizzierten Systeme erlaubt. Ein auf UML aufbauender Graph-Transformations-Ansatz wird genutzt, um die spontane Bildung eng kooperierender Teilsysteme beliebiger Größe zu spezifizieren. Differentialgleichungen beschreiben das kombinierte Verhalten auf physikalischer Ebene. In Kombination ergeben diese beiden Formalismen hybride Graph-Transformations-Systeme, in denen die Graph-Transformationen diskrete Schritte und die Differentialgleichungen das kontinuierliche, physikalische Verhalten des Systems beschreiben. Zusätzlich, präsentieren wir die Erweiterung einer automatischen Analysetechnik zur Verifikation induktiver Invarianten, die bereits für zeitbehaftete Systeme bekannt ist, auf den ausdrucksstärkeren Fall der hybriden Modelle.