A selection of available master thesis projects at SINTEF Digital

This project aims to develop the methodology for controlling the cast shadow of an object, by computing the internal structure of the object, and using novel manufacturing techniques like 3D printing to realize it.

Large simulations are time consuming and cannot be applied real-time in control settings. Instead, we develop machine learning-based surrogate models with relevant information for modelling the controlled environment.

To enable use of machine learning in industry, we need to combine machine learning and physical knowledge. In SFI NorwAI, we offer projects in collaboration with industrial partners on topics such as predictive maintenance for wind turbines and...

This master thesis is carried out in collaboration with the Norwegian Centre on AI for Decisions and potentially with co-supervision from industry partners.

Machine learning has shown success in many different tasks such as image analysis and language generation. However, to apply machine learning to industrial data, we still need to solve challenges with small data, noise, irregular samples etc. This...

Can we use AI to make research communication on websites dynamic and engaging without losing correctness and trustworthiness?

This thesis may cover topics including autonomous navigation and collision avoidance, onboard AI for closed-loop control of data capture, Deep Learning with multimodal sensor data, 3D reconstruction as well as fault detection and condition estimation...

In this thesis you will be given access to real underwater images/video along with CAD models for selected objects and evaluate and demonstrate 6D pose estimation algorithms.

Design and implementation of an inductive wireless-power system for a miniature rodent brain implant, integrating cage transmitter and on-board receiver coil.

The project aims to advance the development of a next-generation diagnostic device capable of rapidly detecting bacteria directly from patient samples - such as blood, urine, or swabs - without the need for time-consuming laboratory cultures.

Development of a microplasma chip as part of a portable breath analyser for real-time, in-home health monitoring within the framework of the large EU EIC Pathfinder project BreathSense

The goal of the thesis is to (1) explore how Large Language Models (LLMs) and LexAlign can be used to assess GDPR and EU AI Act compliance for Telenor’s anomaly‑detection pipeline; and (2) extend LexAlign to support company-specific or sector...

This topic investigates the integration of AI agents in environmental DTs, where we may focus on existing DT-developments such as for the Digital Twin of the Oslo fjord, emphasising their role in data integration, semantic enrichment, and adaptive...

The goal of this thesis is to explore the use of symbolic AI (e.g., knowledge graphs) to enhance reasoning capabilities of GPT and reduce hallucinations and opaqueness, and improve trustworthiness.

The main objective of this thesis is to build and maintain a platform for "Tracking AI Innovation and hype in companies".

The goal of the thesis is to explore the applicability of ChatGPT to enable a higher degree of automation for data integration in real national and European industrial/research projects.

Handling the complexity of Asynchronous Federated Learning in cyber-physical systems.

Ensuring that complex data pipelines adhere to a growing web of legal, ethical, and organizational requirements (e.g., GDPR, AI Act) is a major challenge. This thesis will examine how multi-agent LLMs can be used for the automated compliance checking...

The goal of this thesis is to systematically review the current state of the art in the area of Causal AI, with a focus on causal inference in general, and causal representation learning in particular

The aim of this thesis is to devise a technique for data collection, enrichment, integration, storage, and access for data in the area of green energy production focusing on solar panel data.

This thesis will investigate the use of multi-agent Large Language Models (LLMs) for automating the creation of data pipelines. The work will explore how specialized LLM agents—responsible for tasks such as data ingestion, transformation, validation...

Making Federated Learning even more decentralised to achieve better fault tolerance and resilience.

An exploration of methods and data primitives for enhancing the health of public discourse and governance of media & communities on social media (MIT-led research)

A model-based approach to augmenting AI agents with semantic DISARM for detecting coordinated cyber, media, and human activities in high-stakes elections. (MIT-led research)

An AI-driven analysis of Russian propaganda on social media. Focus: Human-AI teaming, data pipelines, and feature engineering to study identity impact and virality (MIT-led research)

The aim of this thesis project is two-fold: (a) reduce implementation costs of building digital system models using deep learning and semantic technologies, and (b) ensure digital systems models enable domain experts (e.g., engineers and researchers...

We will investigate the way to model and monitor the entire DevOps systems, including the team, the environments, the processes and the final products, using the digital twin concept.

This topic explores the architecture, implementation, and impact of AI agents in asset-oriented DTs, focusing on adaptability, decision-making, and workflow automation.

The goal of the thesis is to explore the use of Large Language Models (LLMs) to enhance data harmonization related to Knowledge Graphs (KGs).

The goal of this thesis is to explore and implement ensemble learning techniques to create a robust system for identifying fish species across different underwater environments, using existing models trained for specific conditions.

This thesis seeks to improve and conduct an in-depth analysis of Entity Linking (EL) problems in public bodies' data sourced from Spend Network.

The goal is to identify and extract entities from the Knowledge Graph that best match or relate to the input text, a crucial process for fields like information extraction, question answering systems, and semantic search [Avogadro et al 2022, Gillick...

The goal of this thesis is to investigate and prototype a system where an LLM assists a data scientist with common time-series data challenges.

The choice between a centralized architecture and a federated one is challenging and often depends on the priorities of the specific application. A centralized architecture seems to be the best solution for guaranteeing accountability and control...

This thesis endeavors to explore software engineering techniques tailored for deploying AI models on devices like the Raspberry PI 5 and SEEED studio from Nordic semiconductor. The research will integrate concepts from intermittent computing and...

This thesis aims to explore the potential of Generative AI in the realm of software engineering, focusing on constructing machine learning pipelines. By synthesizing software components through AI, this research hopes to optimize the design...

This thesis will delve into orchestrating the balance between an edge device (Raspberry Pi) and a more robust system (Desktop or Cloud) using container technologies to realize the benefits of Green AI across the edge-cloud continuum.

This thesis seeks to explore the design and implementation of Guardrails for Generative AI to restrict its output, combined with iterative re-prompting mechanisms, ensuring the generated content aligns with predefined constraints, thus fostering...

This thesis will delve into implementing dataset distillation specifically for time series data sourced from diverse sectors, including health, energy, space, and manufacturing, aiming to pave the way for more sustainable AI applications.

The main goal of the thesis is to build an “Innovation Knowledge Graph” by joining several independent streaming data-sources constituting the main phases of the innovation lifecycle. The aim is to interconnect, model and understand the relevant...

The objective of this work is to propose a model using machine learning (ML) techniques to identify the different data types of columns in a tabular data source. This means establishing a priori categories: PERSON, LOCATION, OTHER, NUMBER, DATE...

This MSc thesis aims at developing an approach based on an existing modern Digital Twin (DT) combines the simulation of the physical twin and the continuous collection of real-time data from the physical twin for real-time monitoring of the systems...

Big data pipelines play a crucial role in various domains that heavily rely on handling vast amounts of data, such as digital healthcare, telecommunications, and manufacturing.

This MSc thesis aims at conducting a systematic review on the approaches (state of the art, SotA) for the Business Continuity of Critical Infrastructures.

This MSc thesis aims at developing an approach for enabling access control of AI/ML applications on Edge to data sources and how the ML/AI-processed data can be shared with different stakeholders as a service.

This topic concerns the design and development of solutions to transform specifications of data enrichment pipelines into executable pipelines.

This thesis seeks to explore the use of Reinforcement Learning from Human Feedback (RLHF) and pretraining techniques to fine-tune small language models for domain-specific applications. Focusing on data from health, manufacturing, and space sectors...

In social sciences, triangulation is a methodology used to validate the reliability of results by cross-verifying them from multiple sources. This thesis proposes a novel approach by applying the idea of triangulation to the outputs of multiple Large...

In this thesis, you will explore approaches for exploiting manual input from the citizens to improve data curation.

Using Geospatial Foundation Models for Species Distribution Modelling

Did you know that developing a drug or a vaccine (which really is a type of drug) takes on the average 13 years and costs up to 3.5 billion US$? And that about 99% of the projects will never make it to the market? And that there are still cases that...

The goal in this thesis is to develop a recommender system to suggest transformations with respect to characteristics of the data set, historical actions users have performed in the past, etc.

In this thesis, you will investigate Machine Learning approaches to improve the curation of structured (SQL databases) and/or semi-structured data sources (e.g., JSON and CSV).

In this thesis, you will explore Machine Learning approaches possibly in combination with heuristic and rule-based approaches for linking entities between data sources.

In this thesis, you are expected to develop a visual query formulation tool for geospatial (i.e., location oriented) data for aiding domain experts to extract data of interest from the data sources encoded in RDF (a data model used for exchanging...

In this thesis, you are expected to develop a visual query formulation tool for temporal data for aiding domain experts to extract data of interest from the data sources encoded in RDF (a data model used for exchanging data on the Web).