Having accurate weather information before taking off on longer journeys can help the user take the optimal routes to reach their destinations depending on the desired goal with the journey. Because of the growing popularity of traveling with a smartphone with an integrated map application, this information can be represented in the form of an application on such a device. The information that is conveyed in the application is integrated from Grönt Väders API, which is a weather forecasting API. For the convenience sake of the user Google Maps API was used to visualize route information on a map. Implementation of  this application is made for the Android platform, hence it’s developed using Kotlin in Android Studio IDE (Integrated development environment). This application was built using  the TDD-method (Test Driven Development) and designed with MVP (Model View Presenter) pattern. The code base is well documented using Dokka (Kotlin version of JavaDoc), manually and automatically tested, and the performance of the application in terms of CPU and Memory usage was measured. While this application was developed for Android exclusively, it is possible to extract the model of the project responsible for reading weather data and with minor changes run the code on a seperate server in order to make the calculations independent of the platform of use and in turn more portable to other devices.

Parkinson's disease (PD) is a progressive, chronic neurodegenerative disorder that impacts patients' quality of life. Hand tremor is a hallmark motor symptom of PD. However, current clinical tremor assessment methods are time-consuming and expensive and may not capture the full extent of tremor fluctuations. The built-in sensors in smartphones offer an accessible and cost-effective alternative for objective tremor assessment. This study presents a systematic approach to developing a quantitative algorithm for Parkinson's disease tremor assessment using Inertial Measurement Unit (IMU) data. This study begins with a comprehensive data visualisation and understanding phase, leading to the design decision to implement a multiple linear regression model for tremor severity prediction. The IMU data, collected from 10 patients, is pre-processed and normalised to ensure consistency and account for varying degrees of tremor severity. Feature extraction is conducted based on insights from literature, resulting in 16 unique features. These unique features are extracted for each of the acceleration and rotation rate data, resulting in 582 total features over both hands and all three tremor types. Recursive Feature Elimination with Cross-Validation (RFECV) is employed for feature selection, identifying the most relevant features contributing to tremor severity prediction. A multiple linear regression model is implemented and trained using the Leave-One-Out with Cross-Validation (LOOCV) method. The model's performance is evaluated resulting in a mean MSE of 0.88, a mean MAE of 0.69, and an R² of 0.88. The results indicate a strong correlation between predicted and actual tremor severity, suggesting the model's high validity. The selected features show a high correlation with the patient's MDS-UPDRS scores, further validating their relevance in predicting tremor severity. Greater results could be achieved, but sample size was the greatest limitation during this study. This study demonstrates the potential of using IMU data and multiple linear regression modelling for accurate PD tremor assessment within Mobistudy, contributing to the field of quantitative PD analysis. / Parkinsons sjukdom (PD) är en progressiv, kronisk neurodegenerativ sjukdom som påverkar patienternas livskvalitet. Handtremor är ett framträdande motoriskt symptom på PD. Dock är nuvarande kliniska tremorbedömningsmetoder tidskrävande och dyra och kanske inte fångar hela omfattningen av tremorfluktuationer. Inbyggda sensorer i smartphones erbjuder ett tillgängligt och kostnadseffektivt alternativ för objektiv tremorbedömning. Denna studie presenterar en systematisk metod för att utveckla en kvantitativ algoritm för bedömning av Parkinsons sjukdomstremor med hjälp av data från Inertial Measurement Unit (IMU). Denna studie börjar med en omfattande data visualisering och förståelsefas, vilket leder till designbeslutet att implementera en multipel linjär regressionsmodell för förutsägelse av tremorseveritet. IMU-data, insamlad från 10 patienter, förbehandlas och normaliseras för att säkerställa konsekvens och ta hänsyn till varierande grad av tremorseveritet. Funktionsextraktion genomförs baserat på insikter från litteraturen, vilket resulterar i 16 unika funktioner. Dessa unika funktioner extraheras för var och en av accelerations- och rotationshastighetsdata, vilket resulterar i totalt 582 funktioner över båda händerna och alla tre tremortyper. Rekursiv funktionseleminering med korsvalidering (RFECV) används för funktionsval, vilket identifierar de mest relevanta funktionerna som bidrar till förutsägelse av tremorseveritet. En multipel linjär regressionsmodell implementeras och tränas med hjälp av Leave-One-Out med Cross-Validation (LOOCV) metoden. Modellens prestanda utvärderas vilket resulterar i ett genomsnittligt MSE på 0,88, ett genomsnittligt MAE på 0,69 och ett R² på 0,88. Resultaten indikerar en stark korrelation mellan förutsagd och faktisk tremorseveritet, vilket tyder på modellens höga validitet. De valda funktionerna visar en hög korrelation med patienternas MDS-UPDRS-poäng, vilket ytterligare validerar deras relevans i förutsägelse av tremorseveritet. Större resultat kunde uppnås, men provstorleken var den största begränsningen under denna studie. Denna studie visar potentialen att använda IMU-data och multipel linjär regressionsmodellering för noggrann PD-tremorbedömning inom Mobistudy, vilket bidrar till fältet för kvantitativ PD-analys.

No description available.

No description available.

Modern education is more and more dependent on technical solutions to assist student learning. Systems specifically designed for this purpose are called learning management systems and provide the link between students and teachers, in the sense that a significant part of their communication is handled within the system. With a learning management system as a core part of a course, the demands on the usefulness and usability of the system are high. The usefulness of an in-house developed system was evaluated in a real world scenario from a teacher’s perspective. The evaluation concludes that the implemented changes increased the usefulness of the system and simplified the work for the teachers.

The times of manual labour are changing as automation grows larger and larger by the day. Self-driving vehicles being one of the more well known examples of automation (the vehicles in this thesis being those found in the construction industry), relies on a machine learning network to recognize its surroundings. To achieve this, the network needs a dataset. A dataset consists of two things, data, which usually come in the form of images, and annotated labels in order for it to learn what it sees. The labels is a descriptor that describes what objects exists in an image, and coordinates for where in the image these objects exists, and the area they occupy. As data is collected, it needs to be manually annotated, which can take several months to finish. With this in mind, is it possible to set up some form of semi-automatic annotation step that does a majority of the work? If so, what techniques can be used to achieve this? How does it compare to a dataset which have been annotated by a human? and is it even worth implementing in the first place? For this research, a dataset was collected where a remote controlled wheel loader approached a stationary dump truck, at various different angles, and during different conditions. Four videos were used in the trainingset, containing 679 images and their respective labels. Two other videos were used for the validationset, consisting of 120 images and their respective labels.  The chosen object detector was YOLOv3, which has a low inference time and high accuracy. This helped with gathering results at a faster rate than what would've been possible if an older version was chosen.  The method which was chosen for doing the automatic annotations was linear interpolation, which was implemented to work in conjunction with the labels of the trainingset to approximate the corresponding values.  The interpolation was done at different frame gaps, a gap of 10 frames, a gap of 20 frames, all the way up to a gap of 60 frames. This was done in order to help locate a sweet spot, where the model had similar performance compared to the manually annotated dataset. The results showed that the fully manually annotated dataset approached a precision value of 0.8, a recall of 0.96, and a mean average precision (mAP) value of 0.95. Some of the models which had interpolated frames between a set gap, achieved similar results in the metrics, where interpolating between every 10th frame, every 20th frame, and every 30th frame, showed the most promise. They all approached precision values of around 0.8, a recall of around 0.94, and an mAP value of around 0.9.

This thesis investigates building a network-on-chip for a multi-core chip computing convolutional neural networks (CNNs) using Imsys processors in a tree architecture. The division of work on a multi-core chip is investigated. Key patterns of communication are identified and three designs allowing for increasingly more advanced communication patterns are implemented in VHDL. Each design is evaluated on throughput, latency and design size by running tests on the communication patterns in simulation. A relation between design size and throughput is shown, though the throughput decreases for different communication patterns when resorting to networks with lower design size. Depending on what layers are present in a CNN of interest, a network can be chosen with as small design size as possible while still achieving desired results. Aspects such as implementation and usage difficulties and energy consumption are discussed in the thesis as well; however, only on a theoretical level.

Mixed, virtual, and augmented reality technology blends virtual and real elements, offering unlimited potential for various industries such as gaming, medical science, and industrial manufacturing. This report aims to show the development process of a system for visualization of and interaction with large and advanced CAD assemblies. The proposed system will enable CAD engineers, service technicians, educators, and other relevant parties to engage with current and emerging products for educational and demonstration purposes without needing a physical model. The report will show the problems and solutions faced when developing such a system and evaluate different approaches. It presents a technical overview of the current state of the mixed reality space concerning the visualization of large and advanced CAD assemblies and what future innovations would be desirable. The result of this paper is a system that imports, converts, optimizes, and adds interaction capabilities of any .STP format file saved on the local file system, all when the software is executing. The system allows for dual-handed interaction of any part, with maintained quality and acceptable performance.

Organizations that do not possess the knowledge or resources to develop IT internally must procure software development projects from external parties. The organization making the procurement is in this thesis referred to as the software customer organization (SCO) and the external party is referred to as the software vendor organization (SVO).  The success of procuring products or services depends on the procurement competence of SCO professionals (Stek & Schiele, 2021). Procurement competence are the individual-level foundations of organisational purchasing and supply management performance (Bals et al., 2019). This thesis aims to investigate what challenges occur within software development projects where SCO professionals have poor procurement competence and build an understanding of how SVO professionals can mitigate or overcome these challenges.  Through semi-structured interviews with project managers from a SVO, six challenges were identified: Fuzzy requirements, Unrealistic expectations, Understanding of the problem, Unaware of the software development process, Inadequate software setting-to-work process, and Poor communication.  Most challenges can be mitigated through a carefully conducted feasibility study beginning by asking questions about the SCO’s previous experiences with software development projects, what competencies they currently have and how they usually work with software development projects. Asking these questions makes it possible to assess the level of the SCO professional’s procurement competence. After assessing, the following actions can be taken: Conducting a workshop on how the SVO’s software development process works, bringing in external requirement experts if needed, and setting up clear communication paths.

Robot vacuum cleaners are used in many domestic and industrial appliances around the world today. The vacuum cleaners have a certain goal: To clean an area in a set amount of time. It does so by applying different techniques based on information from its different sensors.However, the efficiency might vary from robot to robot. The variations are a product of many things as the problem is fairly complex.This paper will measure the variations due to point of entry. In a statistical analysis on a simulated dataset results showed that point of entry does affect the robot vacuum cleaner in some cases. Of the two algorithms simulated in this paper, the random bump algorithm showed little to no benefit at all. Whereas the spiral algorithm showed up to a 20% increase in efficiency due to point of entry. / Robotdammsugare används i många hushåll och industriella tillämpningar runt om i värden idag. Robotdammsugare har ett visst mål: att städa en yta inom en viss tid. Den gör det genom att använda olika tekniker beroende på informationen det får från sina sensorer. Dock kan effektiviteten variera mellan robot och robot. Variationerna beror på många olika saker då problemet är komplext. Denna rapport kommer att mäta variationer som uppstår på grund av startposition. I en statistisk analys på ett genererat dataset visade resultaten att i några fall påverkar startpositionen robotdammsugaren. Med en av de två algoritmerna som simulerades i detta arbete, random bump algoritmen, visade det sig inte spela stor roll. Medans den andra algoritmen, spiralalgoritmen, kunde bli upp till 20% mer effektiv på grund av startposi-tionen.