Cable laying vessels are capital-intensive, long-lived assets whose required capabilities depend on uncertain future offshore wind developments. This thesis investigates how deep uncertainty can be addressed during the early-stage design of a cable laying vessel, focusing on mission-equipment choices such as cable-storage capacity, carousel arrangement, lay-line configuration, and tensioner capacity.

Dynamic Adaptive Policy Pathways is adapted as a design-support framework and combined with a parametric vessel-evaluation model. A library of 11,304 mission-equipment configurations is evaluated against synthetic offshore wind export-cable project portfolios. The scenarios explore increasingly demanding conditions related to cable length, distance from shore, high-voltage direct-current systems, water depth, and floating offshore wind. Performance is assessed using technical feasibility, campaign requirements, fuel consumption, equipment utilisation, and mission-equipment capital expenditure.

The results show that no single configuration performs best under all future conditions. Greater cable-storage capacity improves performance for longer and more remote projects, while higher tension capacity becomes increasingly important in deeper water. Two-line capability is particularly valuable for high-voltage direct-current projects. These insights are translated into adaptive design pathways that identify potential upgrades as future requirements develop. The resulting framework supports transparent comparison of under-specification, over-specification, and adaptability trade-offs before major design decisions are locked in.

The increasing penetration of renewable energy sources (RES), worsening grid congestion and rising electricity price volatility are changing the operating conditions under which prosumers participate in electricity markets. In the Netherlands, these developments are accompanied by increasingly restrictive and time-varying grid connection limits, requiring energy management systems (EMSs) to reduce operational costs while maintaining feasible operation under uncertainty. This thesis develops and evaluates a risk-aware residual reinforcement learning (RRL) framework for day-ahead market (DAM) bidding by a grid-connected prosumer equipped with a commercial building load, a solar photovoltaic (PV) installation and a battery energy storage system (BESS). The proposed method combines a scenario-based optimisation baseline with a learning-based residual policy. This allows the optimisation layer to provide a constraint-aware reference schedule while the reinforcement learning (RL) agent learns corrective bidding and battery-scheduling actions. Time-varying grid limits are incorporated throughout the operational framework, constraining both the day-ahead schedule and the real-time imbalance market operation, where a rule-based controller operates the BESS and curtails PV generation during delivery. The proposed strategies are evaluated under Dutch electricity market conditions and compared against deterministic linear programming (LP) and scenario-based optimisation benchmarks in terms of operational cost, robustness and grid-limit violations.

Building material identification plays an important role in applications such as sustainable building assessment, facade inspection, energy efficiency analysis, and urban environmental studies. Traditional material identification approaches mainly rely on optical imagery or geometric information, which are often limited in capturing the thermophysical characteristics of facade materials. Thermal imagery provides additional information related to heat transfer behavior, emissivity, and thermal inertia, offering an alternative perspective for building material analysis.

This thesis investigates the use of spatio-temporal thermal imagery combined with a Convolutional Long Short-Term Memory (Convolutional Long Short-Term Memory (ConvLSTM))- based deep learning framework for pixel-level building material classification. Groundbased thermal image sequences of building facades were collected under different environmental and temporal conditions using a thermal camera. Building materials including glass, concrete, and brick were manually annotated through Red Green Blue (RGB) imagery and Segment Anything Model (SAM)-assisted segmentation. Image registration techniques were subsequently applied to align the RGB and thermal image sequences, enabling the transfer of material labels into the thermal domain.

A ConvLSTM-based semantic segmentation model was developed to capture both spatial thermal distributions and temporal thermal dynamics from sequential thermal imagery. Different dataset splitting strategies and temporal datasets were evaluated to investigate the influence of spatial generalization and temporal variation on classification performance. The experimental results demonstrate that spatio-temporal thermal sequences provide meaningful discriminative information for building material classification. Under the random sequence-level split strategy, the proposed framework achieved a mean F1-score of approximately 0.686 and a mean Intersection over Union (IoU) of approximately 0.524 across the evaluated material classes.

The results indicate that temporal thermal behavior significantly contributes to distinguishing facade materials. Brick surfaces exhibited relatively stable and spatially uniform thermal characteristics, while glass surfaces demonstrated stronger temporal variation and environmental sensitivity. Concrete exhibited intermediate thermal behavior, making it comparatively more difficult to classify. The findings suggest that incorporating temporal thermal information improves segmentation performance compared with single-frame thermal analysis.

Model-based reinforcement learning can improve sample efficiency by learning a model of the environment and using it for planning. In continuous-control tasks, this allows an agent to evaluate many candidate action sequences before acting. However, planning with a learned model also introduces a failure mode: the planner optimizes a learned return estimate rather than the true environment return, and can therefore exploit errors in the dynamics model, reward model, or value function.
This thesis studies how the model predictive control objective used in TD-MPC-style agents can be made more reliable. The main contribution is EfficientTDMPC, a method that modifies the planner objective by aggregating return estimates across multiple dynamics heads and rollout depths, and by applying disagreement-based pessimism during reanalyze. These changes aim to reduce the variance and exploitability of model-based return estimates while preserving the sample efficiency benefits of latent model-based planning.
EfficientTDMPC is the new state-of-the-art on HumanoidBench-Hard and the hard DeepMind Control Suite (DMC) while matching the state-of-the-art on Easy DMC. The thesis also discusses adaptive horizon selection as a future direction, arguing that planning depth should be treated as part of the uncertainty-aware planning objective. Overall, the thesis shows that the reliability of the learned planning objective is a central design problem in sample-efficient model-based reinforcement learning.

Vulnerability prioritization in financial infrastructure depends on both public vulnerability signals and organization-specific context, such as affected assets, ownership, exposure, compliance status, and remediation constraints. Generic LLM-based systems can help analysts summarize and explain such information, but their direct use in regulated security workflows raises concerns about grounding, traceability, reliability, and control.

This thesis designs and evaluates a context-aware orchestration framework for LLM-supported vulnerability prioritization. The framework separates deterministic enterprise-data retrieval from LLM-based synthesis. For each analyst query, predefined workflows retrieve and join relevant vulnerability, asset, ownership, exposure, and compliance records, prune unnecessary metadata, and validate the generated output before presenting it as analyst-facing decision support.

The framework was evaluated against an autonomous-agent baseline in a controlled offline proof-of-concept environment using an approved static production-derived CSV snapshot from selected operational security systems. The benchmark covered 14 questions across three vulnerability-management use cases: contextual asset triage, risk justification, and remediation guidance. The parametric framework achieved a mean weighted accuracy of 98.4%, compared with 90.0% for the autonomous baseline, with the largest difference in contextual asset triage. It also consumed fewer tokens and showed less degradation under stochastic workflow noise.

The results suggest that constrained, context-aware orchestration can improve grounding, traceability, robustness, and operational efficiency for structured vulnerability-prioritization tasks. The findings should be interpreted as proof-of-concept evidence for the selected benchmark, not as evidence of production-scale deployment performance.