To Örebro University

Background Annually, many persons undergo hip surgery, where urinary retention (UR) is a common complication. A gap exists between clinical practice guidelines (CPGs) and performed care regarding the prevention of UR, highlighting the need for implementation interventions to support uptake of the CPGs. As implementation leaders, first-line managers (FLMs) are key to bridging the knowledge-practice gap and could be included in implementation strategies.

Aim To explore guideline implementation in orthopaedic healthcare with a focus on first-line managers’ implementation leadership.

Method A cluster-randomised hybrid effectiveness/implementation study involving 17 orthopaedic units performing hip surgery at hospitals located in Sweden. Intervention units formed multi-professional internal facilitating (IF) teams (each including at least one FLM) for a 12-month education and support programme to implement bladder monitoring CPGs. Data was collected via interviews (study I, II, IV), medical record reviews (study III), patient surveys (study III), and registry data (study III), and it was analysed using qualitative content analysis (study I) and thematic analysis (study II), statistical analysis (study III), and realist evaluation (study IV).

Results FLMs viewed guideline implementation as a balancing act between organisational structures and unit-level realities (study I). They adapted leadership strategies based on context, often translating guidelines to fit local conditions (study I,II). During COVID-19, implementation was smoother due to unified goals, but it faced barriers such as unclear evidence and rapid changes (study II). The intervention deepened participants’ understanding of both clinical guidelines and implementation science (study IV), though without any statistically significant differences in adherence to bladder monitoring CPGs between intervention and control (study III). Managers played a crucial role in anchoring decisions but stepped back to empower teams (study IV).

Conclusion Although FLMs have a clear vision of beneficial leadership in relation to change and use implementation leadership behaviours in CPG implementation, a support and education programme regarding implementation processes could build on their prior knowledge. Still, leadership actions can be influenced by contextual factors, and the FLMs sometimes have to adjust accordingly. Health organisations could support first-line managers in their important role in reinforcing an evidence-based practice by scrutinising the organisational context and creating processes that support implementation leadership.

The contemporary food system is no longer just a matter of nourishment; it has become a mirror reflecting the ecological, social and psychological tensions of our time. The manner in which we produce and consume food is intricately linked to a number of the most pressing challenges facing society, including climate change and rising health issues. The significance of dietary choices is becoming increasingly evident, with the heart of the matter being why people eat what they do and how people's food choices can be changed. This dissertation explores the role of hedonic and social factors in shaping the adoption of sustainable food choices by exploring three themes: Products, Restaurants and Everyday life. Each theme is explored through a combination of different methodological approaches, including consumer sensory evaluation, field experimentation and a randomised controlled trial enriched with social network analysis.The findings collectively highlight the need for multi-level strategies, which must integrate psychological, social and situational factors in order to be effective. While consumer preferences for the sensory attributes of plant-based meat alternatives differ depending on consumer psychographics (i.e. meat attachment), a preference for products that replicate the sensory experience of conventional meat remains dominant. Beyond sensory drivers, meat also carries symbolic meanings associated with tradition, strength and status. These cultural associations reinforce a 'meat-centric’ gastronomic logic, even in vegetarian formats. However, consumer choice can be influenced when a gastronomic authority (i.e. a chef) recommends a less-popular, non-meat-mimicking vegetarian dish. The same cannot be said for social cues (i.e. cues from other guests), which do not appear to influence consumer choice. It is possible that the fragmentary information – whether factual or social – provided to participants within the studies of this dissertation was insufficiently potent to induce continuous and systematic dietary changes in everyday life, suggesting that influence may stem more from particularly influential individuals than from generic others. All in all, this dissertation provides insights into segmentation strategies that may be considered for future product development within the industry, for use within the food-service sector and for methodological application to future research.

Inflammatory bowel disease (IBD) exhibits significant heterogeneity, encompassing variation in affected gastrointestinal segments and disease course, both within and between subtypes. Proteomics and metabolomics may offer insights into the IBD spectrum and serve as a basis for biomarker discovery, thereby addressing healthcare challenges and advanc-ing our understanding of disease mechanisms. The overall aim of this thesis was to investigate proteomics and metabolomics in relation to IBD.

Patient cohorts are the foundation for many observational studies. In study I, we described the Swedish Inception Cohort in IBD (SIC-IBD), which was used for several subsequent studies. Based on serum proteins, we observed colonic Crohn’s disease (CD) as an intermediate subtype between ileal CD and ulcerative colitis (UC) in study II. In study III, we identified mucosal markers associated with an aggressive course of UC (e.g., MMP-1 and OPG) and developed a mucosal protein signature to predict the disease course. Mucosal angiotensin-converting enzyme 2 (ACE2) levels did not differ between IBD and symptomatic or healthy controls in study IV, but levels were higher in inflamed compared to non-inflamed ileal biopsies from patients with CD. In study V, we found metabolites associated with incident paediatric IBD compared with symp-tomatic controls, among these a positive association for aryl sulfate.

In this thesis, we examined serum proteins, mucosal proteins and exposome-related metabolites in IBD. We identified differences between subtypes of IBD, between inflamed versus non-inflamed tissue, as well as between IBD, symptomatic and healthy controls. These results may inform further biomarker research and support a molecular stratification of IBD.

Task and Motion Planning (TAMP) involves jointly addressing high-level task sequencing and low-level geometric motion planning. While early approaches treated these layers independently, it became clear that tight integration is essential: plans that are valid symbolically may fail geometrically during execution. As autonomous robots are increasingly deployed in complex, unstructured environments, addressing the combined TAMP problem in challenging domains has gained a lot of interest. One such domain gaining attention ove rthe past decade involves granular materials. Granular materials like sand, soil ,and grains exhibit both solid-like and fluid-like behavior, making their interaction with robotic systems highly uncertain and difficult to predict. For instance, the quantity of material a robot can scoop or transport depends on numerous unpredictable factors. This introduces unique challenges for planning and execution, as traditional strategies often struggle to adapt to suc hdynamic conditions. In practice, dealing with granular materials often requires blending pre-computed plans with online feedback mechanisms to remain effective.

To address these challenges, this thesis focuses on the Material Flow Planning Problem (MFPP), namely the problem of efficiently moving granular materials from one place to another. The MFPP lies at the intersection of symbolic reasoning and reliable real-world execution and requires addressing three core tasks: planning actions, navigating to target locations, and executing actions under uncertainty. This thesis proposes an integrated, multi-layered planning and execution framework to bridge abstraction levels, from symbolic to physical execution. Central to this is a formal definition of the MFPP that enables reasoning across symbolic and continuous domains. Based on this, this thesis introduces Athena, a hierarchical planning framework that combines symbolic planning with behavior trees for adaptive and fault-tolerant execution. In order to reach and manipulate the material within these planned tasks, this thesis introduces Navigo, a context-aware navigation system that dynamically selects path planning and control strategies based on real-time feedback from the environment (e.g., terrain, weather, and obstacles), enabling robust autonomous navigation in harsh conditions. Finally, this thesis introduces Atlantis, a modular, multi-layered simulation framework designed to test andiiievaluate planning and control strategies across varying levels of abstraction: from high-level task models to physics-based simulations and real-world trials. Atlantis supports systematic testing and development, helping close the gap between high-level AI planning and low-level robotic execution. Overall, this work offers a robust and adaptable approach to TAMP in domains involving granular materials, providing new methods for planning and execution for complex scenarios both in simulation and real-world applications.

The vision of robots assisting with our daily life tasks hinges on a fundamental challenge: most robot behaviors require programming by experts, creating a barrier for the many. Learning from Demonstration (LfD) is a family of methods with the potential to democratize robot programming by enabling laypeople to teach robots new skills simply by showing them. However, most LfD methods operate as black-box systems, making it difficult for humans to interpret, adapt, or reuse — key factors for human-robot collaboration and understanding.

In this thesis, we instead build upon Behavior Trees (BTs) — a transparent and manageable robot programming framework, whose reactive, modular design and potential for interoperability make them well-suited for constructing sophisticated robot behaviors. To bridge the gap between high-level planning and low-level control, engineers often embed entire behaviors within the BT’s leaf nodes. While functional, this approach encapsulates complete skills, obscuring intermediate subgoals and undermining the transparency and modularity that BTs are meant to provide — ultimately limiting skill reuse and adaptability.

The structure of a BT plays a crucial role in effective behavior design. While several approaches have aimed to learn BT structures from demonstrations, they often rely on predefined action sets and state spaces. This necessity for expert-curated inputs constrains the robot’s learning flexibility and reintroduces a degree of expert dependency. Compounding these challenges is a more foundational issue: the BT community lacks universally accepted definitions and rigorous evaluation methods for key properties such as interpretability and modularity. This absence leads to inconsistent claims and makes meaningful comparisons across different studies a significant challenge.

To address these gaps, this dissertation proposes a path toward more intuitive and effective robot learning, articulated through three key contributions:

First, this thesis formalizes core BT properties for robotics and introduces metrics for the systematic evaluation and comparison of learned policies. Building on this foundation, it investigates how different BT structures impact the interpretability of the control policy, identifying design patterns that best align with human intuition and understanding.

Second, this thesis presents a unified control framework that integrates BTs with a high-frequency Stack-of-Tasks (SoT) control strategy, enhancing transparency of BT policies by explicitly revealing the underlying subgoals. This approach allows BT nodes to function as hierarchical, high-frequency control objectives. Moreover, the resulting system achieves rapid reactivity in dynamic environments while supporting the coexistence of heterogeneous controllers and preserving a clean, modular decomposition of complex tasks.

Finally, this thesis introduces an end-to-end, label-free LfD pipeline that simultaneously learns the global BT structure and the underlying actions —modeled as Dynamic Movement Primitives—directly from raw demonstration data. By leveraging vision-language models to automatically extract and annotate state representations, this method eliminates the need for handcrafted action sets, predefined state spaces, and time-consuming manual labeling.

In summary, this thesis provides a comprehensive framework for learning interpretable, modular, and adaptable robot control policies from demonstration, bridging the gap between transparent policy representation and practical, high-frequency robot control, which marks a significant step toward making robot programming more accessible, robust, and understandable for both experts and non-experts.

Background: Physical exercise (PE) represents a challenge in achieving stable glycemic control in individuals with type 1 diabetes (T1D), especially during prolonged PE. Achieving euglycemia requires careful adjustments of carbohydrate (CHO) intake and insulin doses before, during, and after PE, by proactive use of continuous glucose monitoring (CGM).

Aim: This thesis evaluated strategies for achieving glucose control in T1D before, during, and after prolonged PE, focusing on CHO intake, insulin adjustments assisted by CGM, and education intended to improve self-management.

Methods: Four studies were conducted: (1) High CHO intake with insulin adjustments during prolonged PE; (2) CHO loading, tailored pre-race insulin dose, and high CHO intake during PE; (3) CHO replenishment after evening exercise on nocturnal glucose regulation in healthy individuals; and (4) impact of a diabetes sports camp for adults integrating education and exercise and long-term self-management.

Results: Before PE: CHO loading with basal insulin adjustment maintained stable glycemia. Tailored pre-race insulin supported euglycemia. During PE: High CHO intake (75–100 g/h) with individualized insulin strategies enabled stable glucose levels. After PE: The timing of CHO intake did not affect nocturnal glucose levels when daily energy and CHO needs were met. Self-management: The diabetes sports camp was associated with improvements in perceived diabetes self-management skills.

Conclusion: Glycemic stability in T1D before, during, and after prolonged PE is achievable through individualized CHO and insulin strategies, supported by CGM and structured education.

The increasing complexity and unpredictability of manipulation tasks in modern industrial and service robotics have highlighted the limitations of pre-programmed robot solutions. Robots operating under changing object positions, variable obstacles, and unforeseen perturbations must adjust their actions online to reliably satisfy the high-performance requirements of real-world deployment scenarios.

A promising direction for enabling such flexible and reliable manipulation lies in the use of Behavior Trees (BTs), a formalism for transparent decision-making that structures robot behavior hierarchically through modular, reusable components. BTs are a well-suited solution because their inherent reactivity allows the system to respond effectively to high-level disturbances, such as perception or grasping failures. At the same time, their modular design facilitates the reuse of sub-behaviors across different scenarios, enabling automation systems to be easily reconfigured to meet varying operational demands. However, existing BT-based approaches fall short in scenarios in which more advanced forms of robustness to local perturbations and task variations are required. This thesis contributes novel solutions to address these limitations and enhance the applicability of BTs as control policies in real-world manipulation settings.

To rigorously assess the solutions proposed in this thesis, we first need to formalize the terminology and evaluation criteria associated with BT-based robot control. We begin by identifying a subset of properties that are most relevant to our scope, such as reactivity, modularity, and robustness, and clarifying their definitions by resolving ambiguities found in prior work. For each of these properties, we examine how they have been evaluated in the literature and propose additional metrics to address identified gaps in existing evaluation practices.

The first technical contribution addresses the reactivity of BT policies and their way of handling simultaneous control objectives. While BTs effectivel ymanage global, high-level disturbances, flexible manipulation also requires rapid response to local, low-level perturbations that do not warrant changes to the high-level plan. Furthermore, when BTs are coupled with convention allow-level controllers for redundant manipulators, they often struggle to satisfy multiple, potentially competing objectives in a coherent and reliable manner. To address these limitations, we integrate BTs with a prioritized control strategy that decomposes each manipulation skill, such as grasping, into multiple control objectives with defined priorities, distributed across the BT nodes and executed concurrently. This integration introduces an additional layer of low-level reactivity, ensures the reliable satisfaction of multiple objectives, and reinforces the modularity of the BT policy by assigning distinct goals to separate leaf nodes.

Although the proposed framework provides robustness to both high- and low-level disturbances during execution, it still relies on manually specified parameters, which often need adjustment to specific task variations, such as minor changes of object positions or obstacle configurations. The second technical contribution is a data-driven approach based on Reinforcement Learning that augments the BT with a context-based adaptation policy. This module observes task-relevant features, referred to as a context, and selects appropriate BT parameters at execution time. The result is a policy that adapts its behavior on the fly to previously unseen variations, without manual intervention.

Despite its benefits, the proposed framework remains limited to adapting only to directly observable task variations and requires training procedures that, when performed on a physical robot, are often unsafe and timeconsuming. The last technical contribution addresses both limitations by introducing a context estimator that infers latent dynamics parameters — such as friction coefficients or object mass — from recent interaction data. Conditioning the context-based adaptation policy on this latent estimate enables the BT-based policy to operate robustly even under partial observability. Moreover, because these latent parameters often underlie the discrepancies between simulation and reality, the very same mechanism also provides a principled way to bridge the sim-to-real gap: policies are trained in simulation with domain randomization, while the estimated context compensates for the mismatched dynamics, improving robustness at deployment.

Autonomous systems are increasingly being deployed across various real-world domains, such as fleets of self-driving vehicles, robotic warehouses, and delivery services using unmanned aerial vehicles. These systems are required to operate with high reliability and predictability, to adapt continuously to changing conditions, and to remain accountable to human supervisors. To achieve these objectives, autonomous systems need explicit, formal representations of their behavior that facilitate task planning, system verification, and human oversight.

In current industrial practice, such representations, whether for task-level control or action planning, are typically engineered manually. While handcrafted representations can be precise, their development is labor-intensive and difficult to scale. Learning-based approaches offer a promising alternative by extracting behavioral representations from execution data. However, they often make unrealistic assumptions, such as access to simulated environments or large volumes of high-quality training data. Moreover, they fail to simultaneously achieve all the three critical objectives, that is reliability, adaptability, and interpretability. Therefore, there is a clear need for methods capable of efficiently learning accurate, interpretable representations under realistic conditions.

In this thesis, we address the problem of learning interpretable representations of system behavior from execution traces - sequences of observed actions and state transitions generated during the operation of autonomous systems. Learning from such traces is appealing because they are readily available from system logs and provide direct evidence of how a system behaves in realistic, often complex environments. The overarching goal is to derive representations that not only support automated planning but also enhance human understanding and oversight.

Two distinct types of behavior representation are explored: Behavior Trees (BTs) and STRIPS-style planning domains. For each, a novel method to automatically construct representations from execution traces is proposed. Specifically, for BTs, we introduce a method that combines Boolean logic, leveraging algorithms originally developed in circuit theory, with decision tree learning to induce structured, interpretable behavior representations. To assess the interpretability of BTs, a user study is conducted to examine how such representation sare perceived by human users. The study identifies key features that influence user comprehension, contributing empirical evidence to a space that has traditionally lacked systematic analysis. Furthermore, a structured evaluation method for BTs along with quality metrics and design principles is presented, addressing the current lack of guidance for assessing BT quality beyond functional performance.

For STRIPS-style domains, we introduce a novel learning framework to construct symbolic action representations directly from execution traces, even in the presence of noise. In addition to the learning algorithm, a systematic methodology is proposed for evaluating learned planning domains through structural and task-based analysis, thereby addressing a critical gap in current practice and thus responding to the growing need for rigorous assessment methods.

The results demonstrate that it is possible to extract interpretable representations of autonomous behavior from noisy data. The proposed methods enable the transition from raw execution traces to structured representations that can support planning, validation, and human-in-the-loop systems. By advancing methods for learning, interpreting, and evaluating learned behavior representations, this work contributes to the development of autonomous systems that are both operationally effective and intelligible to human stakeholders.