of Doctoral Thesis Title: Components and Services in Resource-Constrained Envi- ronments Author: Tomáš Pop tomas.pop@d3s.mff.cuni.cz Department: Department of Distributed and Dependable Systems Faculty of Mathematics and Physics Charles University in Prague Advisor: Prof. František Plášil plasil@d3s.mff.cuni.cz Abstract Appliances of every day use such as consumer electronics, automotive and telecommunication devices as well as various kinds of control systems have be- come a common and important part of our everyday lives. In comparison to general-purpose desktop systems, they are limited in terms of various resources, for example memory, CPU power, and battery capacity. Component-based soft- ware engineering is a well-established development technique, which has found its way to industry and has been successfully used for a long time in multiple do- mains. However, the industrially adopted component frameworks were designed for general-purpose systems and they can be hardly used in resource-constrained environments. In the thesis, we explore the potential of component-based system engineer- ing in resource-constrained environments. Stemming from a survey of existing component frameworks targeted to this domain, we identify the most important approaches to overcome resource constraints in...

Softwaresysteme sind ein wichtiger Bestandteil in der heutigen digitalen Welt und müssen oft als fundamentales Investitionsgut von Unternehmen betrachtet werden. Deswegen benötigen solche Systeme eine betriebswirtschaftliche Sichtweise: der Wert als operatives System sowie als langfristige finanzielle Investition bezüglich zukünftiger Anforderungen. Die Grundlage jeden guten Softwaresystems liegt in dessen Architektur, dem Zusammenhang von allen Teilen und deren Beziehungen untereinander, dem inneren Aufbau und der Organisation der Teile. Der Zusammenhang zwischen guter Architektur und resultierender guter Evolutionsfähigkeit und Zukunftsfähigkeit eines Systems ist generell in den Meinungen vertreten. Jedoch liegen dazu nur wenige Literaturstellen vor, welche die Thematik tatsächlich mehr als nur empirisch aufzeigen. Diese Arbeit soll einen möglichen Weg aufzeigen, wie mithilfe eines modellbasierten Ansatzes ein kausaler und quantitiver Zusammenhang zwischen Architektur eines Softwaresystems und dem resultierenden betriebswirtschaftlichen Wert (im Sinne der Evolutionsfähigkeit) hergestellt werden kann.

Toute institution concernée par le développement de logiciels, qu’il s’agisse d’un éditeur ou d’un organisme de services, doit faire face au challenge de compétitivité : “ faire de l’économie, un «objet de première classe» en génie logiciel ”. Cet impératif de penser l’économie du logiciel, s’est traduit par l’objectif d’obtenir une meilleure maîtrise sur la productivité du développeur. La Réutilisation Logicielle (Software Reuse) est un moyen privilégié d’augmenter cette productivité, en particulier quand elle est systématisée. Deux types d’activités doivent être considérés pour améliorer la réutilisation logicielle, le développement pour la réutilisation (development for reuse) et le développement par la réutilisation (development by reuse). Plusieurs solutions ont été proposées pour contribuer au développement pour la réutilisation. L’approche lignes de produits se distingue par sa contribution au développement par la réutilisation, à travers l’assistance et l’automatisation de la sélection, la configuration, et la dérivation de nouveaux produits. Néanmoins, malgré que cette approche a positionné la réutilisation comme activité centrale dans son processus de développement, celle-ci reste difficile à réaliser dans beaucoup de situations. Par exemple, en raison de l’absence de spécification ou gestion de la variabilité qui peut se manifester dans chacun des artefacts de toutes les étapes du cycle de développement. Dans ce cadre, la problématique générale de cette thèse CIFRE consiste en l’industrialisation d’une usine logicielle par la systématisation de la réutilisation dans chacune de ses étapes et la maximisation de l’automatisation des transitions entre étapes. Afin de mieux supporter l’agilité des environnements métier, notre premier objectif est la spécification de la variabilité au sein des modèles de processus métier, de sorte que les modèles conçus soient directement utilisables dans les usines logicielles. Notre second objectif consiste à introduire la spécification de la variabilité lors de l'ingénierie des exigences permettant ainsi de systématiser leur réutilisation, et l’établissement des liens de traçabilité avec les modèles de processus métier variables précédents. Ainsi, un modèle d’architecture logicielle (orientée services) peut être généré au sein de l'usine logicielle, comme implémentation des processus métier modélisés précédemment et dans le respect des exigences spécifiées. / Any organization involved in software engineering has to deal with reduction of production time and cost, in order to face the competitiveness challenge. This imperative of thinking the software economy resulted in the goal of getting better control on developer productivity. Software Reuse is a preferred way to increase the productivity, particularly when it is systematized. Two types of activities should be considered to improve software reuse, development for reuse and development by reuse. Several solutions have been proposed to contribute and improve development for reuse. For its part, product line approach is distinguished by its contribution to development by reuse through support and automation of selection, configuration, and derivation of new products. However, although this approach has positioned reuse as a core activity in its engineering process, it remains difficult to realize it in many situations. For example, due to lack of specification or management of variability which may occur in each artifacts from all steps of the engineering process. In this context, the general issue of this thesis consists in industrialization of software product line, by the contribution to systematization of reuse in each steps and automation of transitions between those steps. To better support the business agility, our first goal is the specification of variability within business process models, in order to make them directly usable into software factory. Our second goal is to introduce variability specification into requirements engineering, enabling systematic reuse of requirements models and establishing traceability links with previous models of variable business processes. Thus, an architecture model (service oriented) can be generated in software factory, as implementation of modeled business processes with compliance to specified requirements.

Service-oriented architecture is a possible solution to the increasing complexity of automotive software systems. We conduct a systematic mapping study investigating the use of service-oriented architecture in automotive software. This study aims at providing information on publication trends, available architectural solutions, main benefits and challenges in automotive service-oriented architecture. From an initial set of 341 peer-reviewed publications, we selected 28 primary studies, which we classified and analysed using a thorough protocol. We provided both quantitative and qualitative analyses using vertical and orthogonal analysis. We found that there has been a significant increase in the number of publications recently, and that the studies focused on defining functionalities and data flow between them. Functionality was found as the most recognized benefit when applying service-oriented architecture in the automotive domain, while security, safety and reliability are the most recognized challenges.

With the substantial growth of cloud computing over the past decade, microservices has gained significant popularity in the industry as a new architectural pattern. It promises a cloud-native architecture that breaks large applications into a collection of small, independent, and distributed packages. Since microservices-based applications are distributed, one of the key challenges when designing an application is the choice of mechanism by which services communicate with each other. There are several approaches for implementing Interprocess communication (IPC) in microservices, and each comes with different advantages and trade-offs. While theoretical and informal comparison exists between them, this thesis has taken an experimental approach to compare and contrast common forms of IPC communications. In this the- sis, IPC methods have been categorized into Synchronous and Asynchronous categories. The Synchronous type consists of REST API and Google gRPC, while the Asynchronous type is using a message broker known as RabbitMQ. Further, a collection of microservices for an e-commerce scenario has been designed and developed using all the three IPC methods. A load test has been executed against each model to obtain quantitative data related to Performance Efficiency, and Availability of every method. Developing the same set of functionalities using different IPC methods has offered a qualitative data related to Scalability, and Complexity of each IPC model. The evaluation of the experiment indicates that, although there is no universal IPC solution that can be applied in all cases, Asynchronous IPC patterns shall be the preferred option when designing the system. Nevertheless, the findings of this work also suggest there exist scenarios where Synchronous patterns can be more suitable. / Med den kraftiga tillväxten av molntjänster under det senaste decenniet har mikrotjänster fått en betydande popularitet i branschen som ett nytt arkitektoniskt mönster. Det erbjuder en moln-baserad arkitektur som delar stora applikationer i en samling små, oberoende och distribuerade paket. Eftersom microservicebaserade applikationer distribueras och körs på olika maskiner, är en av de viktigaste utmaningarna när man utformar en applikation valet av mekanism med vilken tjänster kommunicerar med varandra. Det finns flera metoder för att implementera Interprocess-kommunikation (IPC) i mikrotjänster och var och en har olika fördelar och nackdelar. Medan det finns teoretisk och in- formell jämförelse mellan dem, har denna avhandling tagit ett experimentellt synsätt för att jämföra och kontrastera vanliga former av IPC-kommunikation. I denna avhandling har IPC-metoder kategoriserats i synkrona och asynkrona kategorier. Den synkrona typen består av REST API och Google gRPC, medan asynkron typ använder en meddelandemäklare känd som RabbitMQ. Dessutom har en samling mikroservice för ett e-handelsscenario utformats och utvecklats med alla de tre olika IPC-metoderna. Ett lasttest har utförts mot var- je modell för att erhålla kvantitativa data relaterade till prestandaeffektivitet, och tillgänglighet för varje metod. Att utveckla samma uppsättning funktionaliteter med olika IPC-metoder har erbjudit en kvalitativ data relaterad till skalbarhet och komplexitet för varje IPC-modell. Utvärderingen av experimentet indikerar att även om det inte finns någon universell IPC-lösning som kan tillämpas i alla fall, ska asynkrona IPC-mönster vara det föredragna alternativet vid utformningen av systemet. Ändå tyder resultaten från detta arbete också på att det finns scenarier där synkrona mönster är mer lämpliga.

I världen växer forskning på självgående fordon dagligen.Målet med detta projekt var att skapa ett självgåendefordon och utforska möjligheterna att använda infrarödareflektioner som navigeringsmetod och hur man kanuppnå distinkta mätvärden. Avhandlingen diskuterar ävenmöjligheterna att använda flera prototyper i en störreskala. Under projektets gång byggdes ett prototypfordonför att genomföra experimenten angående lämplighetenmed navigering via infrarött ljus. Tester med prototypenvisar att navigering via infrarött ljus är väldigt pålitligtunder kontrollerade omständigheter. Projektet utforskaräven hur hierarkisk mjukvaruarkitektur står sig mot heltlokal eller centraliserad mjukvaruarkitektur. / In the world, research on autonomous navigation vehicles(AGV) is growing by the day. The goal with this projectwas to create an AGV and explore the possibility of usinginfrared reflections as a navigational method and how toachieve distinct reflection measurements from a surface.The thesis also discusses the possibility of using severalunits on a larger scale. In the progress of the project, aprototype vehicle was built to conduct the experiments toidentify the suitability of infrared navigation. The testingof the prototype shows that navigation by IR can be veryreliable under controlled circumstances. The project alsoexplored how hierarchical software architecture stands incomparison to purely local or centralized software architecture.

In the future, the usage of UAVs (Unmanned Aerial Vehicles) will be applied in many different areas. Continued research in UAVs can benefit through its applications in disaster response, humanitarian aid, environmental monitoring, infrastructure inspection, improved transportation and delivery systems, and scientific research. By leveraging UAV technology and making the technology easier to get into, we can enhance efficiency, safety, and accessibility in various fields while addressing critical global challenges and improving the well-being of individuals and communities. The entry knowledge needed for starting development with UAVs can be high due to the complex communication needed between a ground station and the UAV. This report has the goal of lowering that entry barrier by doing an investigation of some available communication protocols, choosing a protocol and using said protocol to develop and implement an interface for communication between a ground control station and companion computer onboard an UAV. The investigation of communication protocols for UAVs indicated that the widely used Mavlink protocol is the best suited communication protocol for this implementation. The development and iterative process, carried out with the mentioned protocol and a UAV flight controller, resulted in the creation of an artifact that can serve as the desired interface.

L’architecture en microservices met en évidence de multiples avantages pour les entreprises et les développeurs. Cela explique pourquoi de nombreuses entreprises technologiques choisissent de migrer leurs logiciels d’une architecture monolithique vers celle des microservices. Cependant, la migration d’un système monolithique ou d’un système hérité vers une architecture en microservices est une tâche complexe, risquée et chronophage. Pour faciliter et améliorer ce processus, notre travail se concentrera sur la conception d’une approche semiautomatique pour détecter et identifier les microservices dans les applications existantes. Il s’agit d’une étape clé vers l’objectif global de migrer un système monolithique vers des microservices. Notre approche consiste à combiner des méthodes de regroupement et d’optimisation de la proximité entre les dépendances structurelles à un niveau de granularité optimal. Dans ce projet, nous nous appuyons également sur plusieurs méthodes d’intelligence artificielle, en particulier des algorithmes d’apprentissage automatique, pour mettre en oeuvre notre approche. D’une part, nous effectuons l’extraction de dépendances et le regroupement. D’autre part, nous mettons en oeuvre des méthodes qui nous aideront à optimiser la proximité entre éléments constituant un microservice. Pour obtenir les scores de proximité, nous ciblons à la fois les relations sémantiques et les dépendances structurelles. L’analyse des graphes d’appels et des traces d’exécutions peut nous aider à générer les différentes connexions structurelles. En ce qui concerne les connexions sémantiques, nous pouvons tirer parti de techniques d’apprentissage de représentations numériques (embedding) telles que SBERT. / The microservices architecture highlights multiple benefits for companies and developers. This explains the reason why numerous tech companies choose to migrate their software from a monolithic architecture to one of microservices. However, migrating from a monolithic or a legacy system to a microservices architecture is a complex, risky, and time-consuming task. To ease and improve this process, our work will focus on designing a semi-automatic approach to detect and identify microservices in existing applications. This is a key step toward the overall goal of migrating a monolithic system toward microservices. Our approach consists in combining methods of clustering and optimization of proximity between structural dependencies at an optimal level of granularity. In this project, we rely on several Artificial Intelligence techniques as well, specifically Machine Learning algorithms, to implement our approach. On one hand, we are performing dependency extraction and clustering. On the other hand, we are implementing methods that will help us optimize the proximity. To obtain the proximity scores, we are targeting both semantic relationships and structural dependencies. Analyzing call graphs and execution traces can help us generate the different structural links or relations. As for the semantic connections, we can take advantage of highly useful embedding such as SBERT.

Thesis (M. Tech. (Information Technology)) -- Central University of technology, Free State, 2013 / Markets for manufactured products are characterized by a fragmentation of the market (with regards to size and time), and by shorter product cycles. This is due to the occurrence of mass customization and globalization. In mass customization, the same basic products are manufactured for a broad market, but then consumers are given the liberty to choose the “finishing touches” that go with the product. The areas that manufacturers now compete for are higher quality products, low cost and timely response to market changes. Appropriate business strategies and manufacturing technologies must thus be used to implement these strategic dimensions. The paradigm of Reconfigurable Manufacturing System (RMS) has been introduced to respond to this new market oriented manufacturing environment. The design of RMS allows ease of reconfiguration as it has a modular structure in terms of software and hardware. This allows ease of reconfiguration as a strategy to adapt to changing market demands. Modularity will allow the ability to integrate/remove software/hardware modules without affecting the rest of the system. RMS can therefore be quickly reconfigured according to the production requirements of new models, it can be quickly adjusted to exact capacity requirements as the market grows and products change, and it is able to integrate new technology. In this research project, real-time, open controller is designed and developed for Reconfigurable Manufacturing Tools (RMTs). RMTs are the basic building blocks for RMS. Real time and openness of the controllers for RMT would allow firstly, for the modular design of RMTs (so that RMTs can be adapted easily for changing product demands) and secondly, prompt control of RMT for diagnosability.

The software development continues to become more competitive and demanding, placing pressure on developers. Changes in the international political climate have resulted in shrinking military budgets, putting developers of defence software under further pressure. At present, systematic reuse is probably the most realistic way of addressing this pressure by improving software development productivity and quality. Software product line (SPL) engineering provides a comprehensive approach to systematic software reuse and is becoming widely accepted. The focus of this interpretive case study was ground station software development in a small multidisciplinary project-centric company which produces avionics systems for military aircraft. The purpose of the study was to investigate the potential implementation of systematic software reuse in the company. The study consisted of three phases, a literature study, a contextualisation and a set of field interviews, and used elements of the Carnegie-Mellon Software Engineering Institute (SEI) Product Line Practice Framework to examine the suitability of SPL engineering for the company. The findings of the study highlight the potential challenges that SPL engineering poses for the company, and emphasise how the company's project-centric structure could impede its implementation of systematic software reuse. / Computing / M.Sc. (Information Systems)