https://www.mdu.se/

The modern manufacturing paradigm is characterised by an increased level of competition, growing demand for customisable or one-of-a-kind products, and stricter sustainability requirements. To maintain their competitiveness, manufacturing companies must adapt their processes frequently and efficiently while providing high-quality products. Given the importance of establishing flexible and reconfigurable systems, different advanced manufacturing technologies, such as industrial robotics, have seen a drastic increase in usage. However, no system is perfect or free from uncertainties (defects). To achieve Zero-Defect Manufacturing (ZDM, i.e., no defective products leave the manufacturing system, four strategies ‘detect’, ‘predict’, ‘prevent’, and ‘repair’ are needed. However, traditional quality methods, such as quality inspection (detect), suffer from significant limitations in highly customised small batch production.

The objective of this thesis is to facilitate the design of robotic in-line quality inspection systems for ZDM. To achieve the objective, this thesis follows a mixed methods research approach, and its foundation is based on two extensive systematic literature reviews and four case studies in close collaboration with manufacturing companies to investigate how robotic in-line quality inspection is perceived and used. This thesis contributes to the research area of quality management.

Through its findings, this research revealed the unexplicit and partial usage of the ZDM principles in research studies. Thus, this thesis characterises robotic in-line quality inspection, identifies its challenges, and pinpoints its enablers. Robotic in-line quality inspection systems are characterised as ‘connected’, ‘fast’, ‘accurate’, ‘reliable’, ‘holistic’, ‘flexible’, and ‘intelligent’. Several challenges to performing robotic in-line quality inspection have been encountered during this research. As part of the control system, as well as the manufacturing system, performance is highly dependent on its integration with ‘people’, ‘processes’, and ‘technologies’. For example, people need certain competences, time, communication, and participation in the development of ZDM; processes such as ZDM standards are lacking; and available technologies need to be balanced between equipment footprint, interoperability, measurement speed and accuracy, and reliability. Finally, to align all physical, digital, or cognitive components and characteristics, two frameworks and a design flowchart are proposed to help practitioners establish ZDM.

Deep Neural Networks (DNNs) are widely adopted to solve different problems ranging from speech recognition to image classification. DNNs demand a large amount of processing power, and their implementation on hardware, i.e., FPGA or ASIC, has received much attention. However, it is impossible to implement a DNN on hardware directly from its DNN descriptions, usually in Python language, libraries, and APIs. Therefore, it should be either implemented from scratch at Register Transfer Level (RTL), e.g., in VHDL or Verilog, or be transformed to a lower level implementation. One idea that has been recently considered is converting a DNN to C and then using High-Level Synthesis (HLS) to synthesize it on an FPGA. Nevertheless, there are various aspects to take into consideration during the transformation. In this thesis, we propose a multistage framework, DeepKit, that generates a synthesizable C implementation based on an input DNN architecture in a DNN description (Keras). Then, moving through the stages, various explorations and optimizations are performed with regard to accuracy, latency, resource utilization, and reliability. The framework is also implemented as a toolchain consisting of DeepHLS, AutoDeepHLS, DeepAxe, and DeepFlexiHLS, and results are provided for DNNs of various types and sizes.

Patients who suffer from life-threatening illness or injury – experiencing conditions such as cardiac arrest, breathing problems, or trauma – are cared for at designated emergency rooms within the emergency department. In the emergency room, the registered nurse is responsible for those who are exposed and vulnerable and have complex needs. In these rooms, the biomedical focus may reinforce a culture that values the medical-technical aspects of nursing. Meeting patients’ fundamental care needs, such as respect, information, and toileting, in a person-centered way seems challenging in emergency rooms. When care is not provided correctly, the consequences for the patient’s health can be serious, for instance resulting in physical complications in the form of pressure injuries from breathing masks and spine boards or psychological complications such as worry, anxiety, or post-traumatic stress syndrome. Little is known about how person-centered fundamental care is made visible and valued both for and by patients in emergency rooms. In this thesis the understanding of fundamental care is guided by the Fundamentals of Care framework, in order  to maintain an optimal person-centered care that considers the patient’s fundamental care needs with a holistic view of the patient. The overall aim of this thesis is to explore how person-centered fundamental care needs are met for life-threateningly ill patients in emergency rooms, from both patient and registered nurse perspectives.

Study I explored how fundamental care needs of life-threateningly ill or injured patients were met by observing the daily activities of registered nurses in the emergency room, through 108 observations. The results showed that registered nurses were initially engaged and active in meeting patients’ needs, but that this decreased over the duration of the care. Registered nurses met the patients’ physical needs to a greater extent than their psychosocial and relational ones. The environment affected the registered nurses’ ability to meet the patients’ fundamental care needs.

To describe fundamental care needs in the emergency room, based on life-threateningly ill patients’ experiences, an interview study (Study II) was conducted with 15 persons who had been cared for in an emergency room. The interviews were analyzed using deductive content analysis based on the Fundamentals of Care framework. The results showed that relationship, timely and personalized information, and existential needs were identified as essential fundamental care needs, which were not (or only partly) met. The physical environment limited patients in having their fundamental care needs met, and they adopted a “patient role” to avoid adding to healthcare professionals’ stress.

Study III described registered nurses’ work approach and prerequisites for meeting life-threateningly ill patients’ care needs from the perspective of a person-centered fundamental care framework, through 14 interviews. The results revealed that registered nurses structure their work approach in meeting patients’ fundamental care needs based on prevailing organizational and personal prerequisites.

In Study IV the content of guidelines governing the registered nurses’ work in the emergency room was investigated. The results revealed that the registered nurses’ work in Swedish emergency rooms was guided by an instrumental and task-oriented approach to care. The guidelines lacked guidance in providing for patients’ fundamental care needs, and did not support the registered nurses in conducting holistic, comprehensive patient assessments and interventions.

The organizational prerequisites contribute to a task-oriented and instrumental way of working, and patients are not having their fundamental care needs fully met. Fundamental care is not being promoted or prioritized, as the organization and responsibilities for providing person-centered fundamental care are unclear, unspecified, and lacking in direction for how it is to be performed – neither the organization nor the culture supports the registered nurses’ work and profession.

Vi tar för givet att det alltid ska finnas el, rent dricksvatten, mat och läkemedel. Många av våra grundläggande behov tillgodoses tack vare produkter som är beroende av industriella styrsystem. Att skyddas dessa system ifrån störningar är följaktligen ytterst viktigt. 

Vi är mitt i ett teknikskifte som brukar kallas "Industri 4.0" och som innebär att framtidens industriella system kommer skilja sig avsevärt ifrån dagens. Förän­dringen drivs bland annat av nya krav och förväntningar, exempelvis på ko­rtare tid mellan ide och produktion, möjlighet att anpassa produktionen till snabba marknadsförändringar och tillverkning av individuellt anpassade pro­dukter. Flexibla och skalbara lösningar krävs för att kunna uppfylla dessa krav, till skillnad från dagens system som i allmänhet är utvecklade för massproduk­tion av en specifik produkt. 

Detta påverkar såväl hur produktionssystemen konstrueras som designen av varje ingående komponent. En konsekvens är att metoderna som används för att skydda dagens system måste anpassas och stärkas för att möta framtidens utmaningar. En grundläggande sådan säkerhetsfunktion är behörighetshanter­ing. Nuvarande behörighetshantering är inte tillräckligt flexibel och därmed dåligt anpassad till morgondagens dynamiska system. 

I denna doktorsavhandling undersöks behörighetshantering för framtidens industriella system, med fokus på de dynamiska produktionssystem som behövs för att uppfylla kraven kopplade till Industri 4.0. Med utgångspunkt från en enkätundersökning analyseras dagsläget. Förslag på flera olika tillvägagångssätt för dynamisk behörighetshantering presenteras och utvärderas, såväl med avseende på hur sådana regler kan formuleras som på hur de ska kunna upprätthållas. 

Clouds are large computer centers that offer remote access to computing and storage resources, making them popular for business and web applications. They are now being considered for use in safety-critical applications such as factories, but lack sufficient time predictability, which makes it challenging to use them in these time-sensitive applications. To overcome this limitation, an intermediate layer, the fog layer, is introduced to provide computational resources closer to the network edge. However, this new computing paradigm faces its own challenges in resource management, scalability, and reliability due to resource constrained nodes. Lightweight virtualization technologies like containerization can solve the performance-reliability dichotomy in fog computing and provide built-in fault tolerance mechanisms. By studying a robotic use-case, we realized the critical importance of persistent data storage for stateful applications, such as many control applications. However, container-based solutions lack fault-tolerant persistent storage. In this thesis, we identify new challenges associated with leveraging container-based architectures, particularly the importance of persistent storage for stateful applications. We investigate the design possibilities for persistent fault-tolerant storage and propose a solution adapted to container-based fog architectures and tailored for stateful applications. The solution provides scalability, auto recovery, and re-integration after failures at application and node levels. Key elements are a replicated data structure and a storage container, using a consensus protocol for distributed data consistency and fault tolerance in case of node failures. The fault tolerance and consistency of the solution are modeled and verified, and its timing requirements evaluated. We use simulation to evaluate the timing performance of our solution in larger set-ups. The results of our study show that although adding a consistency protocol introduces a timing overhead, the solution still meets timing requirements for the studied use-case even in presence of a set of relevant faults. By leveraging a four-dimensional approach, we also conduct a comparative analysis of our solution with other approaches from various perspectives, indicating that our solution can be applied in a broader context than initially intended.

In recent decades, research has stressed the importance of productive classroom discussions in high-quality mathematical instruction for all learners. However, after decades of attempts by professional development programs to support teachers in achieving these discussions, they are rarely reported on in the classroom context. Thus, there is still a need for support for teachers’ establishment of these discussions. One promising activity for achieving productive discussions involves using a classroom response system and implementing tasks in a multiple-choice format. However, there is little knowledge about whole-class discussions in secondary school mathematics using this approach. This thesis responds to this lack by contributing to knowledge about support for achieving productive whole-class discussions in mathematics using classroom response systems and multiple-choice tasks. The thesis particularly addresses knowledge about the key practices of constructing multiple-choice tasks and leading whole-class discussions. This is operationalized by applying an educational design research approach to 1) develop design principles for constructing tasks and a supplementary task type framework and further evaluate the potential of these design principles by characterizing whole-class discussions, and 2) explore frameworks for analyzing teachers’ leading of whole-class discussions and technology integration. In this work, data from interviews with teachers, reflection notes, and observations were used. The main results cover: characteristics of useful tasks in a multiple-choice format aiming at supporting teachers in achieving productive discussions; characteristics of whole-class discussions in the context of classroom response system in secondary school mathematics; the potential of applying the Structuring Features of Classroom Practice framework as an analytic tool to conceptualize and analyze teachers’ reasoning about classroom response system; and the suggestion of five new categories of teacher actions and enrichments of two existing categories in the redirecting, progressing and focusing framework. In the kappa of this thesis, these results are merged and discussed in relation to theories of ambitious mathematics instruction. The focus is on examining whether and how classroom response systems and multiple-choice tasks can support teachers in establishing this instruction. In summary, the main contributions of this thesis are: a tool for constructing multiple-choice tasks aimed at generating productive discussions in mathematics; an examination of analytical tools for analyzing teachers’ leading of whole-class discussions and technology integration; and suggestions for the development of these tools.

This thesis is an examination of how digital automation of administrative work unfolds in practice. It sets out to understand how administrative work changes as it is digitally automated and how such changes have wider consequences beyond the performance of specific work tasks. A case study design is used, focusing on digital automation through Robotic Process Automation (RPA) at a Swedish municipality, and the methods to produce data include interviews, observations, and document analysis. The thesis contributes to the body of literature that understands work as practices performed by diverse configurations of social and material elements, a body of literature that spans the fields of organization studies and information systems research. It comprises five papers:Paper I builds a foundation for the thesis by examining the automation process and conceptualizing it as configuring work. This is a dynamic process of mutual reconfiguration of work practice, digital technology, and organizational arrangements through which a new agentive configuration of work is approached. Paper II explores the ways in which a new dichotomy of human and digital coworkers emerges and the role of social responsibility and context for work as a new division of labor emerges. Paper III takes a broader look at the effects of digital technology on the organizing of work and proposes the conceptualization of hyper-taylorization as a way of understanding how the rationale of digital automation technology comes to enhance Taylorism in terms of making work digitally legible, predictable, and controllable. Paper IV shifts the focus again to the ethics of digital automation, utilizing an example from the case study to explore ethical and managerial implications when digitally automating. Paper V is a conceptual paper that aims to conceptualize the thesis's core theoretical contribution, which is to understand digital automation of administrative work as not just a change in how work is performed but a change regarding how knowledge about work is created and the conditions of knowledge creation. Within this framework, “work” is understood as performing an epistemic machineryrelated to the materiality of the configuration that performs work. Thus, The paper concludes that digital automation, at least in technological history, implies an epistemological shift of administrative work towards a more strictly rationalistic way of understanding the world at the expense of a pluralistic set of ways of creating knowledge and understanding the world.The thesis concludes by discussing the implications of this shift and how the political terrain of administrative work comes to be abandoned as it is digitally automated.

Battery storage has emerged as a promising solution in various energy systems. However, challenges exist regarding the viability of batteries in practical stationary applications. Factors such as the capital and operational costs, relatively short lifetime, and battery degradation are among crucial factors which have significant impact on battery profitability. To make batteries more viable technology, effective battery management is a necessity. However, there are multiple critical factors which need to be addressed for effective battery utilization and management in real-life applications under dynamic operational conditions.

In this thesis, different battery modelling approaches within battery operational management are proposed. Each proposed scenario consists of a set of specific methods for the estimation of battery performance, capacity degradation, remaining useful life, state-of-charge, state-of-health, and state-of- power.Moreover, the study explores strategies for efficient battery utilization to maximize sustained profitability. Accordingly, the study deals with 32 different state-of-charge operating control strategies as well as different charge/discharge rates (low, moderate, high) to evaluate their impact on techno-economic profitability of a battery system in a grid-connected residential application. Moreover, two day-ahead and optimization-based operation scheduling strategies to maximize battery profitability are proposed. Each scenario employs unique approaches to make optimal decisions for optimal battery utilization. The first scenario aims to optimize short-term profitability by prioritizing revenue gains. Conversely, the second scenario proposes a smart strategy capable of making intelligent decisions on a wide range of decision-variables to simultaneously maximize daily revenue and minimize daily degradation costs.

The key findings reveal that overlooking or simplifying assumptions about multiple critical aspects of battery behavior led to an improper battery management system in practical applications under dynamic operational conditions. Selecting a proper state-of-charge control strategy positively affects the profitability in which alteration of the allowable SOC window from (40%–90%) to (10%–60%) increase the battery lifetime from 10.2 years to 14 years leading to 31.6% improvement in net present value. The key findings showcase how a smart battery scheduling strategy that strike optimal balance between revenue and degradation achieves impressive profit (18-20 €/kWh/year), short payback (7.5 years), and extended lifespan (12.5 years), contrasting revenue-focused scenarios, ensuring sustained profitability for battery owners in residential applications. The findings offer valuable insights for decision-makers, enabling informed strategic choices and profitable solutions.

With the continuous growth of global energy demand, processes from power generation to electronics cooling become vitally important. The role of heat transfer in these processes is crucial, facilitating effective monitoring, control, and optimisation. Therefore, advancements and understanding of heat transfer directly correlate to system performance, lifespan, safety, and cost-effectiveness, and they serve as key components in addressing the world's increasing energy needs.

The field of heat transfer faces the challenge of needing intensive studies while retaining fast computational speeds to allow for system optimisation. While advancements in computational power are significant, current numerical models lack in handling complex physical problems such as ill-posed. The domain of heat transfer is rapidly evolving, driven by a wealth of data from experimental measurements and numerical simulations. This data influx presents an opportunity for machine learning techniques, which can be used to harness meaningful insights about the underlying physics.

Therefore, the current thesis aims to the explore machine learning methods concerning heat transfer problems. More precisely, the study looks into advanced algorithms such as deep, convolutional, and physics-informed neural networks to tackle two types of heat transfer: subcooled boiling and convective heat transfer. The thesis further addresses the effective use of data through transfer learning and optimal sensor placement when available data is sparse, to learn the system behaviour. This technique reduces the need for extensive datasets and allows models to be trained more efficiently. An additional aspect of this thesis revolves around developing robust machine learning models. Therefore, significant efforts have been directed towards accounting for the uncertainty present in the model, which can further illuminate the model’s behaviour. This thesis shows the machine learning model's ability for accurate prediction. It offers insights into various parameters and handles uncertainties and ill-posed problems. The study emphasises machine learning's role in optimising heat transfer processes. The findings highlight the potential of synergistic application between traditional methodologies and machine learning models. These synergies can significantly enhance the design of systems, leading to greater efficiency.