As robots are increasingly integrated into large and dynamic environments alongside humans, there is a pressing need for efficient onboard solutions to fundamental robotic operations, such as navigation and decision-making. Existing solutions often rely on computationally intensive processes that do not scale well in larger environments, leading to long computation times. This can result in unsafe and non-adaptive behaviors, as during the planning phase the robot continues to move along an old and potentially dangerous path increasing the risks of accidents or emergency.This thesis addresses this challenge by developing human-inspired light-weight methods, that enhance robotic navigation and environment understanding.

The central framework presented in this thesis introduces a novel human-like method for navigation and environment segmentation using 2D grid maps, focusing on extracting structural-semantics, such as intersections, pathways, dead ends, and paths to unexplored areas. The framework also generates sparse topometric maps for lightweight robotic navigation by using structural-semantic information. Compared to the state-of-the-art, where map segmentation either utilizes features that are specific to some indoor environments or segments into arbitrary regions that do not convey semantically meaningful information about the environment, the semantic topometric map captures structural-semantic information, which can easily be utilized by robots in a variety of missions. The proposed framework has been validated on multiple maps of different sizes and types of environments. In comparison with the state-of-the-art topological maps generated by Voronoi-based solutions, the proposed framework shows a significant reduction in complexity and computation times required in solving navigation problems. 

The utility of structural semantics is demonstrated through a novel autonomous exploration strategy that integrates structural-semantic information with conventional metric data for goal/frontier selection and employs the semantic topometric map for navigating to a frontier. The effectiveness of the exploration strategy is demonstrated in real-world experiments, showcasing improved exploration speed and computational efficiency compared to frontier-based exploration methods using only metric information. 

In order to enable the methods presented in this thesis to operate over 3D maps, this thesis introduces an approach for converting 3D voxel maps into 2D occupancy maps augmented with height and slope information. Moreover, a method for converting paths generated in 2D into 3D paths is proposed. This allows for the use of structural-semantic segmentation and efficient topometric map-based navigation planning for both UAVs and UGVs. These contributions together enable lightweight and fast environment segmentation and navigation planning for a multitude of robot types, and leveraging structural-semantic information leads to a more human-like approach toward robotic navigation and environment understanding. 

Structural Health Monitoring (SHM) using ultrasonic guided waves is an advanced and efficient technique for evaluating the integrity of critical infrastructure such as pipelines, pressure vessels, and aerospace structures. Despite its numerous advantages, the reliability and effectiveness of guided wave-based SHM systems are often compromised by environmental and operational variations, particularly temperature fluctuations. Temperature changes significantly influence wave propagation by altering material properties, wave velocity, and signal behaviour. These variations lead to signal misalignment and reduced defect detection accuracy, posing a major challenge for real-time monitoring systems. This research addresses the impact of temperature-induced signal variations by developing robust baseline signal strategies to improve temperature compensation in guided wave-based SHM systems.

The study introduces Dynamic Time Warping (DTW) and its advanced adaptations as key methods for aligning signals under fluctuating thermal conditions. DTW is a widely used signal processing technique that optimizes the alignment between two signals by stretching or compressing their time domains, ensuring accurate comparison and analysis. The quadratic computational complexity of DTW poses a significant challenge for its application in Structural Health Monitoring (SHM) using guided waves, particularly for real-time operations under changing Environmental and Operational Conditions (EOCs). 

To address this, the first approach employs a DTW approximation to accelerate the warping process, called Fast-DTW achieving significant computational efficiency without compromising much on performance. In contrast, the second approach uses the Full DTW algorithm but improves its efficiency by incorporating the Sakoe-Chiba constraint to restrict the warping path to a narrow region around the diagonal, effectively reducing computational cost. Additionally, the method leverages the fact that temperature-induced shifts in guided wave signals are typically limited to a few time stamps. Third, we propose a global constraint for temperature compensation of guided waves, referred to as the Triangular Global Constraint (Tri-DTW). The performance of the proposed method is compared with the Sakoe-Chiba Global Constraint (SC-DTW). Tri-DTW performs well, demonstrating better warping performance with four times reduced computational complexity. Finally, we present Hilbert-DTW, a novel approach that extracts the envelope and phase of signals and leverages phase differences to achieve accurate alignment. This method reformulates the cost matrix to prioritize signal alignment while preserving amplitude variations, eliminating the need for constraints or modified DTW variants.

Collectively, these strategies provide a robust framework for baseline signal management and temperature compensation in guided wave-based SHM systems. By improving signal alignment accuracy and reducing distortion caused by temperature effects, the proposed methods significantly enhance the reliability, adaptability, and performance of SHM technologies.

In conclusion, this research advances the field of SHM by addressing temperature-induced challenges through innovative signal processing techniques. The integration of fast DTW algorithms, optimized constraints, and the Hilbert-DTW provides a comprehensive solution for maintaining SHM system performance under varying thermal conditions, enabling more reliable and accurate monitoring of critical infrastructure.

Rotating machinery forms the backbone of energy conversion in modern society, enabling the transfer of energy between a fluid and a rotating element, a rotor.  Mathematical modeling of rotating machinery has become increasingly sophisticated due to the growing demands in modern machine design, primarily driven by the increase of angular momentum sustained by the rotor. Thus, the study of rotating machine dynamics has evolved to improve prediction accuracy and ensure operational stability and performance.

A rotor typically consists of shafts, disks, and blades, supported by bearings that enable rotation relative to a stationary stator. Bearings provide both support and energy dissipation. While rotor-bearing models often suffice for dynamic analysis of rotating machinery, stators—such as casings, housings, and foundations—can influence dynamics through rotor-stator interactions. For example, compliant bearing supports may impair bearing function, while generator stators exert significant electromagnetic forces that impact operational stability. The integrated model of these components is known as a rotor-bearing-stator (RBS) system. 

While traditional rotor dynamic analysis typically utilizes beam elements to model the rotor assembly and reduces the stator and bearing to discrete spring elements. A full three-dimensional representation overcomes the kinematic and geometric restrictions of beam theory, enabling the analysis of vastly more complex systems. Furthermore, the increased degrees of freedom lead to greater computational complexity, particularly in vertical machinery. Unlike horizontal rotors, vertical systems inherently experience large dynamic displacements and lack a stationary point of operation, necessitating the use of nonlinear bearing formulations. In large hydropower machinery supported by journal bearings, this requires tracking rotor position at each time step and solving equations for fluid-film interaction, further increasing computational demands.

This thesis investigates dynamics of integrated systems, with focus on how to accurately and efficiently model complex, vertical systems by finite element methods. The thesis constitutes three applied studies of vertical RBS-systems, mostly related to hydropower applications.

The first paper focuses on a three-dimensional model of a floating-rim synchronous generator, with emphasis on deformation under electromagnetic and centrifugal loads. The second paper explores the interaction between tilting pad journal bearings and an elastic support frame, exploring the potential implementation of nonlinear journal bearing formulations within a three-dimensional finite element framework. Finally, the third paper investigates the impact of a compliant generator-bearing support on the stability of a hydropower rotor-bearing system.

Hydropower is an important renewable energy source providing flexible operating conditions that balances the increasing intermittent energy from wind- and solar power. Hydropower however, comes with other environmental challenges by altering the ecological conditions in the rivers by presenting flow conditions such as hydropeaking. To fulfil the Swedish Environmental Code and the demand from the European Water Framework directive to achieve good ecological status (GES) for all watercourses, measures to improve the ecology in the river for each hydropower plant must be presented. Investigating potential restoration measures might be both costly and time-consuming as they are often specific to the site and species involved. Hydraulic modelling of flow conditions connected with individual-based models (IBMs) to evaluate fish populations are one way to make these estimations more cost- and time-efficient. The aim of this thesis is to investigate how the results of hydraulic models can be used with IBM’s and how different assumptions could create deviations that could further propagate into the IBM’s. A 2D hydraulic model over a regulated river in northern Sweden is presented where the restoration focus is improving European grayling habitat.

As a first step, paper A investigates how modelling parameters could influence the hydrodynamics in the river, and in turn impact the habitat estimations. The results shows that the Neumann boundary conditions are more sensitive to Manning roughness than fixed water level boundaries, especially near upstream and downstream enclosure areas. A fixed water level boundary with velocity measurements at the inlet and outlet can minimize boundary effects on roughness calibration. Careful selection of roughness distribution areas, particularly in high-velocity zones, is crucial as it impacts bed shear stress, and velocities, which can be used for estimating feasible habitat. A gradual roughness transition between calibration areas may improve model accuracy. 

In a second step, paper B compares a steady-state interpolation method with transient simulations since the IBM implements steady-state simulations of different constant flows to interpolate the depth and velocities across the river for different discharge hydrographs. The goal was to assess how the limitations impact different parameters in the IBM’s and how this could affect the long-term habitat. Steady-state and transient simulations were compared, focusing on WSE, spawning habitat and bed shear stress. Steady-state models failed to capture temporal dynamics caused by neglect of time displacement and damping by assuming uniform response times across the river. The bed shear stress was under-predicted by the steady-state interpolation which could lead to inaccurate estimations of suitable spawning grounds and also, the risk of redd scouring. Uncertainties in the depths and velocities over time from the hydraulic model could propagate further in the IBM’s causing the long-term evaluation of fish population to be inaccurate especially when the flow conditions are highly varying during the years where a transient model is preferred. 

Finally, paper C investigates the possibilities of using bed shear stress results for defining the substrate composition in the river. When substrate data is unavailable, bed shear stress can serve as a substitute for identifying suitable habitat areas as input for the IBM. The hydraulic conditions in the river were simulated during a period of 3 months between May and July to capture the spring flood together with some hydropeaking events during summer. The spawning grounds considering depths and velocities was presented with the maximum bed shear stress during the time period. The distribution of possible substrate sizes was presented, providing an indication of the likely substrate composition in the river. The results can be used further to define areas in the IBM but also to identify areas prone to restoration of different substrate sizes and investigate if added substrate would erode due to flow conditions. 

Reinforced concrete is among the most widely used construction materials, yet its durability is often compromised by corrosion. In particular, chloride-induced corrosion poses a severe threat to the longevity of coastal structures. Timely assessment of corrosion levels and monitoring physicochemical changes during the corrosion process are crucial for developing effective anti-corrosion strategies and ensuring structural safety.

Distributed fiber optic sensing (DFOS) has significant potential for monitoring key parameters such as strain, bond stress, and slip in reinforced concrete structures during corrosion progression. Complementary to this, microphysical characterization techniques—including X-ray diffraction (XRD), scanning electron microscopy (SEM), and nanoindentation (DSI)—are indispensable for identifying the composition and structural features of corrosion products and analyzing the micro-mechanical properties of materials. These methods also provide a detailed evaluation of the differences between natural and accelerated corrosion processes. At the nanoscale, molecular dynamics (MD) simulations offer insights into structural weaknesses and enable predictions of corrosion-induced effects. Integrating these multiscale approaches allows researchers to gain a comprehensive understanding of corrosion-induced bond deterioration in reinforced concrete.

This paper begins with a thorough literature review to establish a foundation for the study. It encompasses:

1.      Statistical analysis of MD applications: Identifying MD-based approaches on multiscale studies. 

2.      Key microphysical characterization methods: Highlighting XRD, SEM, and DSI as vital tools for bridging multiple scales.

3.      Macroscopic experimental studies: Investigating bond deterioration in reinforced concrete (RC) tie members exposed to accelerated corrosion, emphasizing DFOS and digital image correlation (DIC) as critical methods.

To facilitate the execution of subsequent formal experiments:

1.      Preliminary tests were conducted to validate the processes for specimen preparation, accelerated corrosion, uniaxial tensile testing, and data collection of RC ties. A pathway for microphysical characterization was also proposed.

2.      At the nanoscale, data extraction methods were optimized, stress-strain curves were corrected, and mechanical property predictions were refined, which are favorable to enhance the reliability of multiscale comparisons and the credibility of experimental results.

Finally, insights from these preliminary experiments and nanoscale model refinements informed the development of a multiscale framework for future studies. This framework integrates macroscopic accelerated corrosion and tensile testing, microphysical characterization, and nanoscale MD simulations, providing a holistic cross-scale assessment of corrosion-induced bond deterioration in reinforced concrete.