The aim of this project was to compare two control algorithms designed to steer an autonomous vehicle. The comparison was made using a simulated environment to evaluate the performance of both controllers. The simulation used in this project was designed in Python and used an algorithm which randomly constructed roads from predefined road segments to create paths for the vehicle to follow. In this environment the Linear Time Varying (LTV)-Model Predictive Controller (MPC) and Pure Pursuit Controller (PPC) algorithms were evaluated. The thesis compared how well they follow paths, the average control cost of completing tasks, how well they handle input constraints, and the computational time for each algorithm. The data was collected by driving along three sets of randomly generated roads with both control algorithms. One set mostly straight, one with some turns, and one with mostly turns. An Analysis of Variance (ANOVA) test was used to make the comparison between the performance of the two algorithms. The results showed that both algorithms performed well. The PPC had low computation time and used less control, but it also had larger position errors. The LTV-MPC had higher computation time, but smaller position errors at the cost of larger control values. The conclusion is that the MPC is preferable if computational capabilities are available. Room for future work exists in the form of comparing additional controller types for autonomous vehicles and exploring different tuning parameters for the MPC controller. The simulation could also be expanded to more accurately reflect real world conditions. / Målet med detta projekt var att jämföra två kontrollalgoritmer avsedda för att styra en självkörande bil. Jämförelsen gjordes med hjälp av en simulering som utformades i Python. Den använde sig av en algoritm som slumpmässigt satte ihop vägar från förkonstruerade delar för att skapa banor för den självkörande bilen att följa. I denna miljö har vi testat två algoritmer, en LTV-MPC och en PPC. Vi jämförde hur pass väl de följer banor som skall likna riktiga vägar, hur mycket styrning de använder sig av för att bedöma energianvändning, hur väl de förhåller sig till begränsningar på acceleration och styrning, och den beräkningstiden som krävdes för att köra vår algoritm. Datan samlades genom att köra längs med tre grupper av slumpmässigt genererade vägar med båda kontrollalgoritmerna. En grupp innehöll huvudsakligen raka sträckor, en innehöll en del svängar, och en innehöll mycket svängar. ANOVA-testet användes för att göra jämförelsen mellan resultatet av dessa två algoritmer. Resultatet visade att båda algoritmer presterar väl. PPCn hade låg beräkningstid och mindre styrvärden, men större positionsfel. MPCn hade högre beräkningstid och större styrvärden, men mindre positionsfel. Slutsatsen är att MPCn är att föredra om beräkningsmöjligheterna finns tillgängliga. Det finns utrymme för framtida arbete i form av att jämföra fler kontrollalgoritmer och att utforska fler parameter justeringar för MPCn. Utöver det finns det även utrymme för en simulation som reflekterar verkligheten noggrannare.

We investigate the novel idea of using computer vision to predict future stock price movement, which is performed by training a convolutional neural network (CNN) to detect patterns in images of stock graphs. Subsequently, we create a portfolio strategy based on the CNN stock price predictions to see if these predictions can generate alpha for investors. We apply this method in the Swedish stock market and evaluate the performance of CNN portfolios across two different exchanges and various stock indices segmented by market capitalisation. Our findings show that trading based on CNN predictions can outperform our benchmarks and generate positive alpha. Most of our portfolios generate positive alpha before transaction costs, while one also generates positive alpha after deducting transaction costs. Further, our results demonstrate that CNN models are capable of successfully generalising their trained knowledge, being able to detect information in stock graphs it has never seen before. This suggests that CNN models are not limited to the characteristics present in their training data, indicating that models trained under one set of market conditions can also be effective in a different market scenario. Our resultsfurther strengthen the overall findings of other researchers utilising similar methods as ours.

Gene editing techniques have been emerging rapidly through the years, and with this trend comes the great responsibility of making sure the edits are correct. One way to safeguard against mistakes in the edits is to measure gene editing efficiency. Countagen’s GeneAbacus does just that, it calculates the gene editing efficiency of CRISPR edits. A key aspect of the GeneAbacus workflow involves quantifying DNA microballs captured in fluorescence microscopy images. This thesis delves into novel image analysis pipelines aimed at optimizing this task. Six image processing techniques (Maximum Intensity Projection (MIP), white top hat transform, Contrast Limited Adaptive Histogram Equalisation (CLAHE), edge enhancement filter, Gaussian Blur, and unsharp masking) along with two object segmentation models (Segment Anything (SAM) and SAM for Microscopy (MicroSAM)) were implemented. They underwent evaluation in two stages: first, through an ablation study of the preprocessing techniques, and then by computing R2 values and log-log plot slopes on different datasets. The evaluation resulted in the selection of MicroSAM with white top hat transform, Gaussian blur and unsharp masking, yielding an average slope value of 0.698 and an average R2 value of 0.8724. / Genredigeringstekniker har vuxit fram snabbt genom åren, och med denna trend följer det stora ansvaret att se till att redigeringarna är korrekta. Ett sätt att skydda sig mot misstag i redigeringarna är att mäta effektiviteten i genredigering. Countagens GeneAbacus gör just det, den beräknar genredigeringseffektiviteten för CRISPR-redigeringar. En nyckelaspekt av GeneAbacus arbetsflöde involverar kvantifiering av DNA-mikrobollar som fångats i fluorescensmikroskopibilder. Detta examensarbete fördjupar sig i nya bildanalyspipelines som syftar till att optimera denna uppgift. Sex bildbehandlingstekniker (Maximum Intensity Projection (MIP), white top hat transform, CLAHE, edge enhancement filter, Gaussian Blur och osharp maskning) tillsammans med två objektsegmenteringsmodeller (Segment Anything (SAM) och SAM for Microscopy (MicroSAM)) implementerades. De genomgick utvärdering i två steg: först genom en ablationsstudie av förbehandlingsteknikerna och sedan genom att beräkna R2 värden och log-log-plottlutningar på olika datamängder. Utvärderingen resulterade i valet av MicroSAM med en white top hat transform, Gaussian Blur och osharp maskning, vilket gav ett genomsnittligt lutningvärde på 0,698 och ett genomsnittligt värde på R2 på 0,8724.

Volvo Construction Equipment (VCE) manufactures wheel loaders, articulated haulers, and excavators. By the end of 2030, the company hopes to have reduced the carbon footprint of its machines by 30 %. To increase energy efficiency and productivity, VCE is focused on developing futuristic wheel loaders that are both electric and autonomous. VCE has unveiled its latest autonomous wheel loader prototype called Zeux. This thesis work aims to create a simulation setup that includes a vehicle model of Zeux and a driver model that is optimised for the machine to complete a certain drive/load cycle. This simulation setup will be used to examine the machine’s performance, energy usage, and efficiency and compare it to a conventional machine to determine its advantages and limitations. The new vehicle model was created by modifying a conventional electric machine’s vehicle modeland a new four-wheel steering system was developed. A driver model was developed based on a condition-based decision tree and state machines with unique controllers for each driver input. This complete vehicle-driver simulation set-up has been tunedand optimised with respect to energy efficiency and productivity. The simulation results are then compared to the results of a similar conventional electric machine simulation model. According to the comparison study, the autonomous wheel loader concept has better productivity, lower hydraulic energy consumption as well as lower overall energy consumption compared to the conventional machine. It can complete the drive cycle much more efficiently despite having a similar powertrain and loading unit as the conventional machine. / Volvo Construction Equipment (VCE) tillverkar hjullastare, midjestyrda dumprar och grävmaskiner. I slutet av 2030 hoppas företaget ha minskat koldioxidavtrycket för sina maskiner med 30 %. För att öka energieffektiviteten och produktivitet är VCE fokuserade på att utveckla framtida hjullastare som både är elektriska och autonoma. VCE har presenterat sitt senaste autonoma hjullastarprototyp som heter Zeux. Detta examensarbete syftar till att skapa en simuleringsmiljö som innehåller en fordonsmodell av Zeux och en förarmodell som är optimerad för att maskinen ska klara en viss kör-/lastcykel. De framtagna modellerna ska sedan användas för att undersöka maskinens prestanda, energianvändning och effektivitet och jämföra resultaten med en konventionell elektrisk maskin för att fastställa dess fördelar och begränsningar. Den nya fordonsmodellen skapades genom att modifiera en konventionell elektrisk maskins fordonsmodell och ett nytt fyrhjulsstyrningssystem utvecklades. En förarmodell utvecklades baserad på ett tillståndsbaserat beslutsträd och tillståndsmaskiner med unika regulatorer för varje drivrutin. Den kompletta simuleringsmodellen har justerats och optimerats med avseende på energianvändning och produktivitet. Resultaten jämfördes sedan med simuleringsresultat av en liknande konventionell elektrisk hjullastare. Enligt jämförelsestudien, har konceptet med autonoma hjullastare bättre produktivitet, lägre hydrauliskenergiförbrukning samt lägre total energiförbrukning jämfört med den konventionella maskinen. Den kan slutföra körcykeln mycket mer effektivt samtidigt trots att den har en liknande drivlina och lastenhet som den konventionell maskin.

Semantic segmentation is the process of assigning a specific class label to each pixel in an image. There are multiple areas of use for semantic segmentation of remote sensing images, including climate change studies and urban planning and development. When training a network to perform semantic segmentation in a supervised manner, annotated data is crucial, and annotating satellite images is an expensive and time-consuming task. A resolution to this issue might be self-supervised learning. Training a pretext task on a large unlabeled dataset, and a downstream task on a smaller labeled dataset, could mitigate the need for large amounts of labeled data. In this thesis, the use of self-supervised learning for semantic segmentation of remote sensing data is investigated and compared to the traditional use of supervised pre-training using ImageNet. Two different methods of self-supervised learning are evaluated, a reconstructive method and a contrastive method. Furthermore, whether including modalities unique to remote sensing data yields greater performance for semantic segmentation is investigated. The findings indicate that self-supervised learning with in-domain data shows significant potential. While the performance of models pre-trained using self-supervised learning on remote sensing data, does not surpass that of pre-trained models using supervised learning on ImageNet, it achieves a comparable level. This is notable given the substantially smaller training data used. However, in cases where the in-domain dataset is small — as in this thesis with approximately 20,000 images — leveraging ImageNet for pre-training is preferable. Furthermore, self-supervised learning demonstrates promise as a more effective pre-training approach compared to supervised learning, when both methods are trained on ImageNet. The reconstructive method proves more suitable for semantic segmentation of remote sensing data compared to the contrastive method, and incorporating modalities unique to remote sensing further enhances performance.

Modern deep learning models require large amounts of data to train, and the acquisition of data can be challenging. Synthetic data provides an alternative to manually collecting real data, alleviating problems associated with real data acquisition. For recycling processes, classifying metal scrap piles containing hazardous objects is important, where hazardous objects can be damaging and costly if handled incorrectly. Automatically detecting hazardous objects in metal scrap piles using image classification models requires large amounts of data, and metal scrap piles contain large variations in objects, textures, and lighting. Furthermore, data acquisition can be challenging in the recycling domain, where positive objects can be scarce and manual acquisition setup can be challenging. In this thesis, synthetic images of metal scrap piles in a recycling process are created, intended for training image classification models to detect metal scrap piles containing fire extinguishers or hydraulic cylinders. Synthetic images are created with physically based rendering and domain randomization, rendered with either rasterization or ray tracing engines. Ablation studies are conducted to investigate the effect of using domain randomization. The performance of models trained on purely synthetic datasets is compared to models trained on datasets containing only real images. Furthermore, photorealistic rendering with ray tracing rendering is evaluated by comparing F1 scores between models trained on data sets created with rasterization or ray tracing. The F1 scores show that models trained on purely synthetic data outperform those trained solely on real data when classifying images containing fire extinguishers or hydraulic cylinders. Ablation studies show that domain randomization of textures is beneficial both for the classification of fire extinguishers and for the classification of hydraulic cylinders in metal scrap piles. High dynamic range image lighting randomization does not provide benefits when classifying metal scrap piles containing fire extinguishers, suggesting that other lighting randomization techniques may be more effective. The F1 scores show that synthetically created images using rasterization perform better when classifying metal scrap piles containing fire extinguishers. However, when classifying metal scrap piles containing hydraulic cylinders, images created with ray tracing achieve higher F1 scores. This thesis highlights the potential of synthetic data as an alternative to manually acquiring real data, particularly in domains where data collection is challenging and time-consuming. The results show the effectiveness of domain randomization and physically based rendering techniques in creating realistic and diverse synthetic datasets.

This thesis investigates the application of knowledge transfer techniques to process event-based data forperson detection in area surveillance. A teacher-student model setup is employed, where both modelsare pretrained on conventional visual data. The teacher model processes visual images to generate targetlabels for the student model trained on event-based data, forming the baseline model. Building onthis, the project incorporates feature-based knowledge transfer, specifically transferring features fromthe Feature Pyramid Network (FPN) component of the Faster R-CNN ResNet-50 FPN network. Resultsindicate that response-based knowledge transfer can effectively finetune models for event-based data.However, feature-based knowledge transfer yields mixed results, requiring more refined techniques forconsistent improvement. The study identifies limitations, including the need for a more diverse dataset,improved preprocessing methods, labeling techniques, and refined feature-based knowledge transfermethods. This research bridges the gap between conventional object detection methods and event-baseddata, enhancing the applicability of event cameras in surveillance applications.

Point cloud registration is an extensively studied field in computer vision, featuring a variety of existing methods, all aimed at achieving the common objective of determining a transformation that aligns two point clouds. Methods like the Iterative Closet Point (ICP) and Fast Global Registration (FGR) have shown to work well for many years, but recent work has explored different learning-based approaches, showing promising results. This work compares the performance of two learning-based methods GeoTransformer and RegFormer against three baseline methods ICP point-to-point, ICP point-to-plane, and FGR. The comparison was conducted on data provided by the Swedish Defence Research Agency (FOI), where the data was captured with a photon-counting LiDAR sensor. Findings suggest that while ICP point-to-point and ICP point-to-plane exhibit solid performance, the GeoTransformer demonstrates the potential for superior outcomes. Additionally, the RegFormer and FGR perform worse than the ICP variants and the GeoTransformer.

[pt] Neste trabalho, aborda-se a modelagem e controle de um quadricóptero para navegação autônoma em ambientes agrícolas. Os modelos cinemático e dinâmico do veículo aéreo são computados a partir do formalismo de Newton-Euler, incluindo efeitos aerodinâmicos e características das hélices. O sistema de movimento do quadricóptero pode ser dividido em dois subsistemas, um translacional e outro rotacional, responsáveis pelo controle de posição nos eixos x, y, z, and atitude do veículo no espaço Cartesiano. A primeira abordagem de controle é linear, se presenta dois controladores, um controlador proporcional-derivativo (PD) e o adaptativo baseado no espaço de estados. A segunda abordagem é não-linear e baseada em um controlador adaptativo a fim de lidar com a presença de incertezas nos parâmetros do sistema. Simulações numéricas são executadas em Matlab para ilustrar o desempenho e a viabilidade da metodologia de controle proposta. Simulações computacionais 3D são executadas em Gazebo para verificar a navegação autônoma em um campo agrícola. / [en] In this work, we address the modeling and control design of a quadrotor for autonomous navigation in agricultural environments. The kinematic and dynamic models of the aerial vehicle are derived following the Newton-Euler formalism. The motion system of the quadrotor can be split into two subsystems, that is, translational and rotational subsystems, responsible for controlling the position along the longitudinal, transverse and vertical axes of the Cartesian space as well as its orientation about the corresponding axes. The first linear control approach is based on the proportional-derivative (PD) controller, whereas the second nonlinear control approach is based on an adaptive controller in order to deal with the presence of uncertainties in the system parameters. Numerical simulations are carried out in Matlab to illustrate the performance and feasibility of the proposed control methodology. Gazebo was used to perform the 3D simulations for verifying autonomous navigation in agricultural fields.

Artificial intelligence (AI) is one of the fastest developing technologies globally. AI has recently entered warfare and thus taken a place in international law. Today the use of AI in warfare is through machine learning and autonomous weapon systems. Autonomous weapons are expected to play a decisive role in future war- fare and therefore have a major impact on both civilians and combatants. This gives rise to an examination of the role of artificial intelligence, machine learning and autonomous weapon systems in international law, specifically international humanitarian law (IHL).  The purpose and main research question of the thesis is to examine how the use of AI, machine learning and autonomous weapon systems is regulated within international law. Further the thesis examines if the regulations sufficiently can ensure the protection that is guaranteed within IHL or if additional regulation is needed. The research question is answered by examining the relevant rules in IHL, compliance with the protection stated in the principles of distinction, pro- portionality and precautions in attack and lastly by analyzing the consequences for civilians and combatants.  Conclusions that can be made is that the rules of IHL are both applicable and sufficient to, in theory, regulate autonomous weapon systems. However the weapon itself must be capable to follow IHL and in order to guarantee this ad- ditional regulation is needed on the use of autonomous weapons. The use of autonomous weapon systems does not necessarily violate the principles of dis- tinction, proportionality and precaution in attack. On the contrary, the use of autonomous weapons can possibly ensure that the principles are respected even further. This however depends on the actual capabilities of autonomous weapon systems and whether they can make the complex judgments required for each principle. It is although still of importance to ensure that the element of human control is never completely lost. The issue that keeps returning is the potential loss of human control. At all times human control must be guaranteed to ensure that the final decision always remains with a human. If humanity in warfare is lost the consequences and risks for civilians will increase. Not only is there a possibility of increase in use of violence but also an increase of indiscriminate attacks. The rules of IHL aim to protect civilians as well as combatants, and the use of this new weapon will lead to difficulties to navigate armed situations for combatants. This will increase the suffering of civilians, but also risk overriding the protection of combatants that IHL ensures.