Intraday electricity markets (IDM), which is designed to correct forecast error of renewable energy generations and enable energy trading, are characterized by high volatility and rapid price fluctuations, which not only provide market participants with strong motivation to make accurate price predictions, but also present significant challenges. The use of machine learning methods for price prediction has become a major trend in recent research. However, in previous studies, only a few specific features, such as Volume-Weighted Average Price (VWAP) and last transaction price \cite{abstract167, lasso39, abstract168}, have been applied, while the rich features embedded in orderbooks have not received sufficient attention. Furthermore, while quantile regression tasks, which provide richer information for trading strategies, have been employed in IDM price prediction, they have generally been confined to statistical models \cite{sta131, lasso38}. Deep learning-based quantile regression, capable of capturing nonlinear relationships and incorporating uncertainty, has yet to be applied. Additionally, in current research, IDM price prediction is often based on a specific orderbook, and thus, the generalization of prediction methods across different orderbooks, as well as their structural similarities across different markets and product types, has not been convincingly addressed.

To address the challenges mentioned above, this report focuses on the German and Austrian markets over continuous trading periods from January 2022 to January 2025, considering both hourly and quarter-hourly products. A total of 384 feature candidates were extracted, including percentiles, momentum, and volatility of prices and trading volumes on both buy and sell sides across multiple time windows. For the extracted feature candidates, we propose an innovative feature selection approach based on their correlation with normal and extreme price labels. Comparative experiments demonstrate that this algorithm outperforms L1-based selection and Principal Component Analysis (PCA) compression in quantile forecast evaluations. Based on the selected features, Quantile LightGBM (QLGBM), Quantile Extreme Gradient Boosting (QXGB), Quantile Multilayer Perceptron (QMLP), and Quantile Kolmogorov–Arnold Network (QKAN) were used to predict the labels, providing a comprehensive set of benchmarks. Additionally, generalization studies across markets and products were conducted using transfer learning, with multiple strategies such as zero-shot, fine-tuning, and joint learning applied. The results reveal valuable insights into the relationships between different markets and product types.

Conventional articulated robot arms excel on structured production lines but remain unsafe and ineffective in cluttered, dynamic settings, while existing soft-robotic manipulators sacrifice load-bearing capacity and positional repeatability for safety and adaptability, origami-robots improve this trade-off. The octopus arm, a muscular hydrostat, is a biological example of overcoming this trade-off by uniting shape morphing, distributed actuation and tunable stiffness. This thesis introduces a fold-flat, modular origami structure that co-locates distributed magnetic actuation and thermally tunable stiffness inside each unit module, thereby matching the octopus arm’s three biological features, shape morphing (1), distributed actuation (2) and tunable stiffness (3). A four-legged water-bomb module was created by mapping biologically derived requirements onto an additively manufactured construction that integrates paired NdFeB permanent magnets, a central pancake electromagnet, and a Joule-heated conductive-PLA crease. Physical prototypes and finite-element analysis predicted ±20 ° biaxial bending, 100 % axial extension, and a 50% stiffness reduction. A proof-of-concept prototype verified these predictions, except that the bending was up to 10°. Single-leg or single-hinge tests revealed a steep modulus drop between 45 °C and 60°C, yielding up to 90 % stiffness reduction under 34 V excitation. At module level the actuator stack produced 10 mm axial stroke, ±10 ° bending and peak push/pull forces of 0.8 N/1.0 N while maintaining repeatable 3-DoF motion. Eight of the ten primary requirements were met; sustained horizontal self-support and whole-arm 90 ° curvature were limited by cable weight and hinge play. The study identified two bottlenecks: high coil currents that drive the PLA above its glass transition, and asymmetric flux coupling when the coil drifts toward one magnet. Improvement strategies include closed-loop thermal control, heat-resistant substrate polymers, and improved electromagnet positioning. By uniting shape morphing, distributed actuation, and tunable stiffness in a centimetre-scale, magnetically actuated origami cell, this work realises an octopus-like robotic arm with power-off stiffening, planar manufacturability, and modularity. The results establish a viable route toward deployable continuum arms for confined-space inspection, human-robot collaboration, and search-and-rescue operations.

Underwater detection has garnered increasing attention in recent years due to its broad and impactful applications in marine ecological research, underwater structural inspection, archaeological exploration, and deep-sea resource extraction. However, despite the proliferation of research in this domain, a comprehensive methodology that addresses both object identification and localization in underwater scenarios remains absent. Existing studies tend to treat these two tasks separately, often omitting the practical implementation details necessary for real-world deployment. This fragmented approach limits the effectiveness and adaptability of underwater detection systems, particularly in dynamic or unpredictable marine conditions.
To bridge this gap, this report presents a detailed exploration of a camera-based framework for simultaneous object identification and localization in underwater environments. The proposed system leverages a Region-based Convolutional Neural Network (RCNN) for object identification, offering a favorable trade-off between precision and computational efficiency. RCNN's architecture enables it to effectively handle the complex visual features typically present in underwater imagery. For the localization task, two complementary strategies are employed: the Metric3D depth estimation algorithm, which utilizes learned monocular cues to infer depth maps with high accuracy, and a geometry-based method rooted in camera imaging principles, which estimates object distance based on intrinsic and extrinsic camera parameters. The former localization method (Metric 3D) is more computationally expensive, but it provides more robust results and easier applications.
Experimental evaluations demonstrate that the proposed integrated approach achieves robust performance in various underwater conditions. The RCNN consistently delivers accurate object classifications, while the localization strategies offer flexibility and reliability depending on the computational and environmental constraints.
Overall, this research contributes a novel and practical solution for real-time underwater object detection by unifying identification and localization. The proposed system enables safer navigation, more precise manipulation, and greater situational awareness. By addressing the methodological gaps in existing literature and emphasizing real-world applicability, this work contributes to the research of intelligent underwater operation and automation.

The Nahuel Huapi National Park, in the Lake District of Northern Patagonia, Argentina, is well known for its tourism industry all year round. After COVID-19, the area saw a significant increase in the number of tourists traveling to the area. This means that the lake located in the heart of the district, Lago Nahuel Huapi, is being used more and more to explore the environmental richness of the area by boat. Now, the capacity of mooring spaces is no longer sufficient in the region, resulting in the construction of illegal private docks along the shore. To reduce this impact on the environment the authorities granted in 2024 a concession to develop one of the last not yet commercialized marina’s in the region: the marina in Bahía López.

This report provides a consult for the concessionaire of this development. The process begins with a research phase, consisting of an area study, and the mapping of environmental and hydrodynamic constraints. Subsequently, stakeholders are categorized, as the development of a marina in a national park entails complex regulations from multiple organizations. The outcomes of the research phase are translated into specific functional requirements for the marina. These functional requirements are the basis for the next phase, the design phase. This phase begins with the formulation of a design vision statement, formulating the project response to local conditions. Based on this, three different conceptual designs with various technical solutions are developed. Through a multi-criteria analysis, the concepts are tested on their robustness in order to chose a final concept. This concept is then elaborated into a preliminary design. Presenting an overview of the marina’s facilities, including structural designs, operational needs, and capital costs. Finally, suggestions for future development
are provided, outlining the next steps to advance the marina to a next phase.

This thesis presents the design process of an active wire rope tensioner, a device designed for cranes, that serves two primary functions. Firstly, it increases the tension in the wire rope during spooling, ensuring neat and tight winding on the drum to reduce damage to the wire rope and cutting-in problems. Secondly, it reduces the required lower block weight, thereby improving the crane's load curve.

The research has commenced with a study of relevant background information and working principles. A thorough literature and patent review has followed, revealing a gap in the state-of-the-art of tensioning devices that both prevent cutting-in and reduce the required lower block weight. In the conceptual design phase, seven innovative concepts have been generated based on the literature review and current principles. These concepts have been assessed for their capability to meet the requirements and their performance against the key performance indicators (KPIs). Consequently, five concepts remain, with the clamping track concept emerging as the most promising.

The clamping track concept utilizes a chain drive with clamps attached to it that press on the wire rope. By applying force to the chain, the tension in the wire rope can be manipulated. Detailed development of the clamping track concept has yielded a conceptual design with three potential deployment scenarios, each requiring a slightly different version of the active wire rope tensioner and offering distinct advantages and disadvantages. Scenario 1 offers the greatest possible lower block weight reduction but comes with the cost of a more complex and riskier system. Scenario 3 does not allow for any lower block weight reduction but is simpler and safer. Scenario 2 strikes a balance between the two.

This thesis presents the optimization of chord-bracing connections (tubular joints) in lattice boom structures using Wire Arc Additive Manufacturing (WAAM). The research aims to enhance the static and fatigue performance of critical joints, with a focus on the 1600mt LEC crane boom, which is used for offshore wind turbine installation. Through the integration of topology optimization and WAAM, the study seeks to improve structural efficiency by optimizing weld geometry and material distribution.
Finite Element Analysis (FEA) was employed to identify high-stress regions within the tubular joints, followed by topology optimization to refine their design. The study also investigates the impact of WAAM on material efficiency, fabrication flexibility, and the overall mechanical properties of the joints. The results demonstrate that the optimized tubular joints exhibit significant improvements in fatigue life and static strength compared to traditional designs, providing a more robust and cost-effective solution for lattice boom applications.
The findings of this research contribute to advancing the use of additive manufacturing in structural engineering, particularly in enhancing the performance and sustainability of offshore crane systems. The proposed optimization framework and design methodologies offer valuable insights for future applications of WAAM in heavy-duty structural components.

Fused Deposition Modelling (FDM) is one of the most popular 3D printing technologies because of its affordability and accessibility. FDM, however, often suffers from printing errors that result in wasted time, materials and energy. To address these challenges, this thesis introduces a novel fault detection system for FDM printers. This system is designed to identify a broad range of errors without interrupting the printing process. To achieve a real-time detection system, an innovative multi-camera setup is designed, integrating two side cameras and one nozzle camera. Our hypothesis is that a system including three cameras can provide a more comprehensive view and can ensure more error types to be detected. Error detection is achieved using Convolutional Neural Networks (CNNs). This is a type of machine learning that excels at image recognition and pattern detection, making it well-suited for identifying printing errors in real-time models. Two CNN models are developed to classify images into common 3D printing errors: one model for the nozzle and another for the side cameras. The models were trained and validated on diverse datasets containing various shapes, infills, and augmented data. The nozzle camera model achieved a high validation accuracy of 97.68% with a low loss of 0.07464. The side camera model achieved comparable performance with a validation accuracy of 97.61% and validation loss of 0.1196. These two well performing models were for the first time ever integrated into a unified fault detection system based on a logic-driven priority framework. From this research, we learned that integrating multiple viewpoints into a logic-driven priority framework significantly improved the robustness of error classification, as many more error types could be detected in-situ and real-time. As a result, the integrated system successfully detected 12 common printing errors. In summary, this work shows the feasibility of developing a robust multi-input fault detection system to improve 3D printing. It paves the way for further research and implementation for complex integrated error detection and correction mechanisms.

Global container trade continues to grow while climate policy tightens emission and efficiency requirements in major hub ports. European ports such as Rotterdam must accommodate higher volumes and stricter reliability expectations without proportional expansions of quay length or yard space. Within this context, intra-port inter-terminal container exchanges remain a bottleneck. Inland barges provide a low-emission alternative to trucks, but fragmented planning and long waiting times undermine reliability and the business case for modal shift. This thesis examines whether autonomy-enabled modular splitting of barge calls can improve operational performance at the Maasvlakte container terminals.

A port-wide discrete event simulation is developed with deep sea, feeder, conventional barge, and autonomous module services calling five terminals that share a dedicated module-crane pool. Current multi-stop barge operations are compared with scenarios in which barges detach short-calling modules under different barge–module mixes, sea-freight demand levels, module-crane inventories, and module capacities. Performance is assessed using throughput, turnaround and waiting times, berth occupancy, crane utilisation, and anchorage behaviour.
In the recommended 50-50 barge-module mix, modular splitting increases port-wide throughput by about 12% and reduces barge turnaround by more than half compared with current operations, while deep sea and feeder vessels remain largely unaffected. These gains arise from parallel module calls that use residual quay pockets more effectively. Hybrid fleets with around half of inland work carried by modules therefore provide the best compromise between higher throughput and manageable growth in vessel calls. Configuration experiments indicate that two dedicated module cranes per large terminal are sufficient under the tested loads and that medium-sized modules around 24 to 36 TEU perform robustly. Within the limits of the stylised simulation, the results indicate that modular intra-port services can improve inter-terminal operational performance, support port sustainability and modal-shift objectives, and provide guidance for the design of MAGPIE Demo 6.

Tailored load-displacement mechanisms are a type of nonlinear mechanisms that can be designed to obtain a specific desired load-displacement curve. In order to achieve tailored torque-twist curves, these mechanisms often need to be complex; requiring multiple parts that sum up to the desired curve. In this work a revolute compliant mechanism, based on a helicoidal shell, is adapted so that it can achieve positive tailored torque-twist curves, using a single part. The goal of this paper is to make a tool that allows this tailorable property to be fully realised using the the rate of twist along the spinal axis as the design variable, whilst the radius and thickness are kept constant. This tool has an analytical model as a basis that predicts the numerical result of a Kirchoff-Love shell model. The model is dimensioned by linking numerical results to shape factors of the helicoidal shells as well as observed deflection behaviour. The resulting design method is demonstrated with the creation of a few tailored torque-twist curves: A linearly increasing, as well as a linearly decreasing curve, and two sinusoidal curves, designed to be used as gravity balancers. Subsequently these designs are produced and measured.
Related dataset 4TU.ResearchData: https://doi.org/10.4121/d87be8e0-38bb-4bdd-a703-70c3e856b1ca

This thesis develops a conceptual design methodology for hydroponic systems tailored to specialty crops where biological requirements are often incomplete or uncertain. The proposed methodology adapts established mechanical engineering design principles (Pahl & Beitz, Roozenburg & Eekels, TRIZ) by introducing iterative feedback loops, explicit decision points, and the parallel integration of biological, technical, and economic analyses. A critical innovation is the inclusion of a dedicated Testing phase between the Conceptual and Provisional design phases. The methodology was applied to a case study on Russian dandelions, a potential alternative source of natural rubber that grows in its roots. The case study successfully structured the complex design problem, generated multiple cultivation concepts, and systematically exposed critical knowledge gaps regarding root rubber content, single- or multiple harvesting techniques, and cultivation strategies. Experimental trials confirmed the feasibility of hydroponic cultivation but revealed significant biological challenges, such as plant stress from root trimming. Economic modeling, based on current assumptions, indicated a lack of viability, highlighting a dependency on future agronomic research. The thesis contributes to both literature and practice by bridging engineering methodologies with controlled-environment agriculture, expanding the scope of hydroponics beyond food crops, and offering a design-support guideline for future innovation in non-traditional crop systems.

This thesis investigates vibration transmission during \ac{vp} with a focus on risks to the lifting equipment. The work is motivated by incident reports and by the absence of field measurements on cranes during offshore operations. The study aims to identify resonance frequency(ies), quantify component displacements, determine forces in the hoist cable, and characterize how each component responds across the input frequency range. A secondary aim is to relate frequencies that favor pile penetration to potential risks for the crane.

A two stage modeling approach is adopted. First, the crane is generalized to a \ac{2d} form and converted to a \ac{1d} system of lumped masses and linear springs aligned vertically. The pedestal is idealized as fixed. Linear behavior and small displacements are assumed. Second, the \ac{mp} is modeled in Ansys with shell elements to capture flexible body behavior. A mode reduction retains only axial modes of the \ac{mp}, since bending, torsion, and circumferential shell modes do not directly couple to the vertical vibration path that governs \ac{vp}. Depth dependent stiffness and damping represent the soil at the toe. Hoist stiffness varies with depth through cable length.

The calculation method separates free and forced vibration. In free vibration, the system eigen-problem provides resonance frequency(ies) and mode shapes with the \ac{mp} first treated as rigid. The axial flexible body natural frequency of the \ac{mp} is then obtained from the shell model. In forced vibration, the harmonic response is computed to obtain frequency response functions and absolute displacement magnitudes for the components, as well as the hoist cable force.

Results show two frequency families that govern the response. The first family contains the system resonance frequency(ies) with a rigid \ac{mp}. The second family contains the axial flexible body natural frequency of the \ac{mp}. Modes 1, 2, and 4 are largely insensitive to \ac{mp} flexibility. Mode 3 and the axial flexible body frequency depend strongly on the soil stiffness at the toe and shift with depth. Within the operational range of the \ac{vh}, the fourth system resonance and the first axial flexible body frequency of the \ac{mp} lie close together and can interchange order as depth changes.

The harmonic response clarifies component participation at key frequencies. Near Mode 3, motion concentrates in the exciter and \ac{mp} and engages the soil stiffness most strongly. Near Mode 4, the lower block and bias mass dominate while the \ac{mp} response is limited. At the axial flexible body frequency of the \ac{mp}, the head and toe move in opposite directions with a near stationary point along the pile length, consistent with a fundamental axial mode. These behaviors explain the locations and amplitudes of the observed peaks.

A verification step compares boom tip response and hoist cable force between the simplified system and the full \ac{fe} crane. The first peak aligns in both models, supporting the validity of the simplified representation for global behavior. Differences at higher frequencies are traced to flexible crane substructures that the simplified model does not include. This comparison establishes where the simplified model is reliable and where detailed crane is more suitable for cranes structure fatigue study.

Mitigation is explored conceptually. Removing the hoist load path would eliminate force transmission into the boom, but this is often impractical. Introducing an isolator between the lifting equipment and the piling equipment is a more practical option. When tuned near the axial flexible body frequency of the \ac{mp} and near the fourth system resonance, an isolator can reduce force transmission into the boom while preserving penetration performance.

The study concludes that frequencies that favor penetration can occur near frequencies that amplify responses in the lifting equipment. Managing this close proximity requires attention to frequency modes, awareness of depth effects through soil stiffness, hoist force fluctuations, and consideration of isolation in the load path. The work provides a structured method to identify critical frequencies, quantify responses, and separate the roles of monopile flexibility and crane structure, while pointing to targeted measurements and modeling extensions that would complete a validated framework for decision making.

The global energy sector is increasingly shifting towards sustainable solutions, driving the expansion of offshore wind farm installation to meet rising energy demands. The upcoming wind farms will be constituted of larger wind turbines, requiring larger monopile foundations. The monopiles are typically installed using impact hammers that blow several strikes until the monopile reaches a specific penetration depth. The installation process generates high underwater noise levels, posing severe risks to marine life, including hearing loss, behavioral changes, and even death.

Existing noise mitigation systems struggle to attenuate low-frequency noise below1000 Hz due to the mass law limitations of natural materials. This thesis proposes a design methodology for a novel metamaterial-based noise mitigation system–the meta-cushion–designed to brake this limitation. Placed between the monopile and the hammer, the meta-cushion filters out mechanical waves generated by the hammer’s blow that cause high underwater noise levels. The design methodology ensures adaptability to various monopile installation specifications, facilitating the implementation of the meta-cushion in real world pile driving cases.

The thesis begins with an analysis of underwater noise during pile driving and the limitations of existing mitigation solutions, emphasizing the need for novel approaches to reduce low-frequency noise. The analysis shows a strong relation between high underwater noise levels and the low-frequency vibrations of the monopile caused by the hammer’s blow. Then, the concept of metamaterials is introduced, highlighting how their unique properties contribute to reducing low-frequency vibrations. Based on the understanding of the underwater noise-vibration relationship, an initial meta-cushion design exhibiting such filtering features is defined.

The numerical modeling of the meta-cushion is then detailed, with finite element analyses being conducted to investigate the behavior of the meta-cushion when interacting with mechanical propagating waves. The attenuation capabilities are evaluated via analysis of dispersion-curve and transmission loss diagrams. To ensure both mechanical integrity and noise reduction under impact loads, a design optimization strategy is developed, from which genetic algorithm is used to investigate the trade-off behavior between mechanical and attenuation performances. Once verified, the meta-cushion’s functionality is experimentally validated through modal impact analysis and small-scale pile hammering tests, demonstrating effective noise reduction at frequencies below 1000 Hz as also indicated by the numerical results.

Finally, this thesis presents a systematic design methodology–encompassing meta-cushion design selection based on pile specifications, numerical verification, and experimental validation–that can be adapted to full-scale monopile installations. By integrating the meta-cushion into offshore pile driving operations, this work contributes to mitigating underwater noise and its environmental impact.

This thesis describes an investigation in the modeling and manufacturing of Nitinol (NiTi) superelastic metamaterials. To achieve an accurate agreement between model and experiment regarding superelasticity in Ni-rich NiTi metamaterials, the research focuses particularly on the martensitic transformation behavior. Each chapter contributes to a better understanding of the entire additive manufacturing process chain, including constitutive modeling, heat treatments and structural design, to develop NiTi-based metamaterials with tunable superelastic properties.
NiTi shape memory alloys exhibit unique shape memory effect (SME) and superelasticity due to reversible martensitic transformations. These properties make NiTi a suitable material for adaptive structures, biomedical devices, and aerospace components. Though computational models can be used to design NiTi structures and metamaterials with superelasticity and SME, the successful additive manufacturing of these designs remains challenging. Achieving full superelasticity in complex geometries produced via laser powder bed fusion (L-PBF) is particularly difficult. Thus, aiming for model-experiment consistency of superelastic metamaterials, the main research gap lies in establishing clear qualitative and quantitative relationships between material models, properties, mesoscopic structures and the resulting macroscopic responses.

In Chapter 3, the research started with the development of analytical expressions for the effective transformation stress and numerical models to evaluate the superelastic behavior and energy dissipation of truss-based metamaterials. NiTi truss-based metamaterials with body-centered cubic (BCC) and octet structures were selected to represent bending- and stretching-dominated architectures, respectively. A detailed parametric finite element analysis was performed to study the relationship between relative density and effective transformation criteria. Using L-PBF, crack-free BCC and octet samples were successfully fabricated from Ni-rich NiTi powder. However, the as-fabricated samples exhibited only partial superelasticity and premature fracture.
In Chapter 4, the study is focused on inhomogeneity of microstructural and functional properties and the underlying reasons for partial superelasticity. Through both numerical and experimental approaches, the study investigated how geometric factors, such as relative density, affect microstructural inhomogeneity and thermomechanical properties of NiTi in body-centered cubic (BCC) structures. Geometric effects on melt pool behavior lead to different solidification textures and inhomogeneous response to indentation. The numerical simulation shows that inhomogeneous transformation temperatures cause narrow stress hysteresis in the macroscopic response. This chapter reveals the interdependent relation between relative density, microstructure, localized properties of NiTi and the macroscopic response.

The focus in Chapter 5 is on understanding and mitigating premature fracture in Ni-rich NiTi metamaterials produced by L-PBF. To investigate the origins of fracture, a comparative analysis of two unit cell architectures, the a Gyroid network (bending-dominated) and a Diamond shell (stretching-dominated), was conducted. Due to the inherent tension-compression asymmetry of NiTi, the structural stability of these designs was found to be reversed compared to conventional elastic-plastic responses, leading to premature fracture and limited superelasticity in the as-fabricated samples. As large deformation can not be achieved through martensitic transformation or dislocation slip systems, partial superelasticity and low deformation recoverability were observed in the as-fabricated samples. Heat treatments were applied to address these issues and achieve qualitative agreement between experimental data and model predictions.

After the superelasticity was successfully achieved, the transformation stress-temperature relation and energy absorption in Ni-rich NiTi superelastic metamaterials were investigated in Chapter 6. Temperature dependence often restricts the practical use of superelasticity. In metallic metamaterials, energy absorption typically relies on the elastoplasticity of ductile metals; however, achieving energy absorption with recoverable deformation has not been fully explored. To address this, a numerical model of the Diamond shell structure was developed to predict temperature-dependent superelasticity and energy absorption. A heat treatment was applied to ensure agreement between the model and experimental results. The findings show that the transformation stress-temperature coefficient decreases from 9.5 MPa/°C for bulk samples to 0.9 MPa/°C for Diamond samples. Under uniaxial compression, the effective transformation stress can be controlled by relative density, with values of 41.8, 52.1, and 65.3 MPa for relative densities of 0.15, 0.2, and 0.25, respectively. A specific energy absorption of 3.5 J/g was achieved in cyclic compression tests with 15 cycles. The recoverable plateau-like response in the macroscopic stress-strain curves originated in a continuous transformation region forms along the macroscopic [100] direction under uniaxial compression. Post-yielding plasticity in the macroscopic stress-strain curves is related to a plastic shear band formed along the [110] direction.

In summary, this work successfully developed a model-manufacturing strategy for superelastic NiTi metamaterials. By addressing multiscale challenges such as microstructural inhomogeneity and tension-compression asymmetry, this study demonstrates that computation-based design and additive manufacturing can create functional NiTi structures with tunable thermomechanical properties. Multiscale issues often prevent computational designs from being fully realized in experiments. By identifying and controlling variable interdependencies across scales, this research achieves largely tunable superelasticity in experiments. This approach provides a foundation for practical applications of NiTi metamaterials in fields such as biomedical devices, aerospace, and civil engineering.

This thesis presents a model-based design framework for incorporating mechanical intelligence into soft robotics. By integrating smart materials with morphing structures, the framework enhances the adaptability of soft robotic systems. The embedded mechanical intelligence enables the synchronization of multiple smart materials, facilitating self-actuation and customized deformations. This approach advances the development of autonomous and adaptive soft robotic systems.

This masters thesis investigates the influence of various printing parameters on the shape memory effect of 3D printed objects, with a focus on fixity and recovery rates. Through a series of experimental tests, it was observed that printing temperature had minimal impact on fixity and recovery rates. However, correlations were identified between fixity rate and layer height, as well as between percentage infill and recovery rates. Lengthwise shrinkage, particularly prominent in samples with 0\% infill, was attributed to printing speed the formation of voids within the structure and layer height. Higher printing speeds were found to compromise mechanical properties while facilitating enhanced shape transformation responses. Additionally, changes in layer height led to observable alterations in the printed object's geometry, including bending and bulging, due to retained shape memory of the filaments original form. Moreover, certain 100\% infill samples exhibited an unexpected hardening phenomenon akin to annealing. These findings underscore the intricate interplay of printing parameters in determining shape memory properties, mechanical properties and highlight potential avenues for optimization in 3D printing processes.

Flexible robotics offers promising solutions for navigating complex environments, and this study contributes to its advancement through innovative methodologies optimizing both flexible robotic arm kinematics and a novel flexible actuator. The first methodology focuses on optimizing kinematics using quadratic programming, enabling the determination of optimal segment configurations without prior knowledge of specific materials or working principles, thus introducing a novel systematic first step in flexible robotic design. Addressing computational intensity and solver compatibility limitations, valuable insights are provided into designing segmented flexible robotic arms, offering a systematic methodology for real-world challenges. Simultaneously, an electromagnetically actuated Kresling cylinder is introduced, leveraging tunability in axial stiffness and electromagnetic actuation. Through dimensional optimization using finite element analysis (FEA), critical design considerations such as coil dimensions and core configurations are systematically explored. Experimental validation extends the application to full-sized robotic arms for package unloading in confined spaces, underscoring the significance of magnetic force in overcoming gravitational resistance, especially in logistics environments. These methodologies represent significant contributions to the field of flexible robotics. The first provides a systematic framework for kinematic optimization, while the second introduces an innovative actuation mechanism tailored for flexible robotic arms in industrial settings requiring long-reach capabilities. Their integration opens new possibilities for designing adaptable robotic systems capable of complex tasks in diverse environments. By addressing computational challenges and practical constraints, this research advances the frontier of flexible robotics, facilitating real-world implementation across various industries.

This thesis explores the fabrication of biomimetic shark skin, focusing on actively controlling the bristling effect of denticles, inspired by the passive bristling effect observed in the skin of Shortfin Mako sharks. By examining the natural structure and functionality of shark skin, the research aims to replicate its unique properties using innovative fabrication techniques. A review of previous methods for replicating shark skin structures led to the selection of an indirect fabrication approach, driven by the necessity to integrate an actuation mechanism. Drawing inspiration from soft robotics, the study identified Nitinol, a shape memory alloy, as the optimal material for actuation due to its significant force exertion and seamless integration into the silicone shark skin.
Experiments were conducted in a water tank, testing various configurations of biomimetic shark skin to evaluate the performance of the integrated actuation mechanism. The initial findings demonstrated that Nitinol wire actuation within a silicone matrix effectively mimics the bristling mechanism, even in aquatic environments, showing sufficient strength to counteract water flow. While this research successfully demonstrated the feasibility of integrating actuation mechanisms into biomimetic shark skin, further studies are necessary to optimize the design for practical applications.

In mechanical systems, energy efficiency is often reduced by unwanted forces like friction and backlash, which significantly impact performance and, over time, lead to increased wear and maintenance costs. Implementing compliant joints can address these issues, simplifying assembly and improving performance. However, compliant mechanisms introduce higher internal loads due to elastic deformation in the flexures, limiting the range of motion and affecting overall efficiency. One way to solve this is using the principle of static balancing, which ensures that a point of interest remains in equilibrium, meaning that no added forces or moments are needed to keep it at rest at certain positions. This can be done by compensating the internal potential energy, which has been implemented in literature for 1 Degree of Freedom configurations. The goal of this paper was to expand the known method into designing a multi-degree of freedom without the use of external springs. The method was based on applying a preload to the compliant joints in the mechanism and optimize in such a way to create a constant potential energy curve. The model was validated using finite element analysis, prototyping and physical testing. It can be concluded that for the first time, a two degree of freedom statically balanced mechanism has been designed. It has a peak force reduction of $95\%$ and a moment reduction of $80\%$, while having a range of motion worth $50\%$ of the size of the mechanism. The mechanism can be used in vibration isolators or as fundamental inspiration for new applications where preloading could lead to lower exerted loads and thus improve efficiency.