Chalmers Open Digital Repository
Not a member yet
26247 research outputs found
Sort by
Adaptive Parameter Control for Search-Based Unit Test Generation
No full textTesting is a crucial task in software engineering, but one that is time-consuming and expensive. Test generation frameworks, such as Pynguin, are designed to automate the process of creating tests, thus alleviating some of the cost. However, the test suites generated by Pynguin do not always manage to reach 100% code coverage, thus indicating that there is room for improvement. This thesis aimed to explore whether adding reinforcement learning-based parameter control to update parameter values during the test generation process could yield improved code coverage. We created PynguinAPC, a version of Pynguin with the addition of reinforcement learning-based parameter control. To evaluate the impact of parameter control, we performed experimental simulation in two cycles, one where we applied control to individual parameters and one where we applied control to pairs of two parameters for a set of 24 Python modules. The results were analyzed using Bayesian statistical models, concluding that there was no overall gain for the final branch coverage achieved by the generated test suites and the branch coverage growth rate from the application of parameter control. The addition of a parameter control system resulted in an overall increase in the performance overhead. However, some modules were more receptive to parameter control for specific parameters, warranting an investigation of what traits these modules inhabit that caused this increased receptiveness. Paradoxically, by introducing a system to reduce the amount of necessary manual configuration, the system inadvertently introduced additional layers of configuration. These configurable layers could merit further exploration into their impact, as the choices made for this study may be a limiting factor in the results observed
Research and Development of Data Science and AI Tools for Project Management at Trafikverket
No full textThis master’s thesis explores the application of product development methodology to the design and prototyping of data science tools within the public sector. The work was conducted in collaboration with Trafikverket, Sweden’s national transport administration, to support more efficient project management in large infrastructure projects.
The study investigates how AI and data-driven tools can be identified, evaluated, and developed at a public agency. The research follows a structured product development process comprising user needs analysis, opportunity identification and screening, concept development, prototyping, and deployment assessment. Initial insights were gathered through interviews, meetings, and a survey conducted with Trafikverket staff. The data of which was collected and assembled into a needs list. Several potential AI tool concepts were proposed and screened throughout these conversations. Four high-potential concepts were screened through a ranking and comparison to the needs list.
One of these concepts, an AI-supported clustering tool for managing public consultation feedback (Samråd), was chosen for prototyping. Using Swedish-language text embeddings and the BERTopic framework, the tool was designed to automatically group and summarize incoming insights to reduce manual workload and increase response consistency. This opportunity also underwent a deployment assessment of economic, environmental, and social impacts. Based partly on the research knowledge from the project, but also based on a follow-up interview with a consultant at SWECO. This project demonstrates that utilizing a product development process was not only
appropriate, but could also have been beneficiary as opposed to traditional data science processes for an organization with limited knowledge of their own needs for data science and AI tools and how to implement them. The outcomes serve as a road map and proof of concept for Trafikverket’s continued adoption of AI tool
Building a Motor Test Bench Prototype
No full textIn modern production machines, numerous electrical motors are used, and failures are common. To reduce downtime and improve reliability, anomaly detection by sensors can be used. The anomaly detection algorithms can discover abnormal patterns in the motor’s performance and thus, give an early indication when failure is underway.
An issue with this, however, is that the subject is complicated to study, as real-life production machines are difficult to test with different failure scenarios.
The purpose of this project was to set up a test bench with two DC motors and sensors that could be used for studying anomaly detection algorithms.
A test bench was built with an MDF board as the base. On the board, two 3D-printed motor holders for the DC motors were attached along with two axis holders to mount an axis between the motors. An Arduino Mega was connected to control the motors and adjust the speed with a potentiometer. Additionally, a temperature sensor and a 3-axis accelerometer sensor were applied to the motor and connected to the Arduino Mega.
In conclusion, a prototype of a test bench for further development was created
Designing a Standardized Marking System for Public Service Drones
No full textThe increasing interest in drone innovation for public services, such as emergency medical deliveries and early situational awareness, raises new challenges around public trust and transparency. This thesis, conducted in collaboration with Region Västra Götaland (VGR), investigates how a standardised marking system can help communicate the purpose and origin of public service drones to the general public. By aligning the needs of diverse stakeholders, including VGR, drone companies, and citizens, the study aims to reduce ambiguity and increase acceptance of drones into shared public spaces. Using methods including interviews, questionnaires, and iterative prototyping, the project identifies key concerns around surveillance, recognition, and ethical concerns. The resulting design framework, MarkeD, including visual markings such as colour schemes, patterns, and placements of logos, is intended to signal trustworthiness and to differentiate drones from private or military use, as well as support visibility. The findings offer both a practical framework for drone marking and broader insights into designing for societal acceptance in emerging technologies as well as academic discourse on new and controversial technologies
Dynamic load of timber truck
No full textThis project investigates the dynamic load generated by heavy timber truck combination
as part of a collaborative research effort between Chalmers university of Technology,
NTNU and volvo trucks. The overarching goal of the research is to improve the understanding
of how trucks and their loads influence bridge structure, with the long term
objective of enabling safe reclassification of exiting swedish bridge fie higher load limits
(BK4).
In this project,the focus on developing a simplified yet representative dynamics model
of a timber truck combination that can later be used for studying Dynamic Amplification
factor(DFA) and truck-bridge interaction. The part involved deriving the equation
of motion for the truck and trailer, identifying and estimating key model parameters,
and implementing the model in a suitable simulation environment. Parametric studies
were carried out to analyze how factors such as road surface irregularities, vehicle speed,
suspension characteristic and axle configuration affect the resulting dynamic loads.
Furthermore, the project included the planning and preparation of experimental test
using an instrumented timber truck for future model validation. The developed model and
proposed testing methodology together provide a foundation for accurately simulating the
dynamic behaviour of heavy vehicle and for supporting future studies on the interaction
between vehicle and bridges
Statistical Evaluation of Radar Simulation Models Towards Real Data
No full textThis thesis aimed to create a method that could be used by Volvo Car Corporation
(VCC) to statistically evaluate the truthfulness and accuracy of the built-in
radar models used by VCC. Using pre-collected data from European New Car Assessment
Programme (Euro NCAP) scenarios—specifically Car-to-Car Rear moving
(CCRm) and Car-to-Bicyclist Nearside Adult Obstructed (CBNAO) — a method
was developed to extract real world data, enabling the recreation of actual scenarios
within a simulated environment. Furthermore, a comparison between the simulated
and real data was conducted. This comparison was conducted using the statistical
metric Double Validation Metric (DVM), which is a combination of the Area Validation
Metric (AVM), Corrected Area Validation Metric (CAVM), and the model
bias within the simulated data.
The developed method includes parsing of Hierarchical Data Format version 5
(HDF5) files, enabling reading and manipulating these files. It also features a Graphical
User Interface (GUI) that reads one real file and one simulated file, visualizes
the desired detection columns within a given timestamp interval, and a statistical
evaluator that performs all the calculations, plots the Empirical Cumulative Distribution
Function (eCDF), and analyzes their statistical characteristics.
The results demonstrated tendencies pointing towards both reliability and unreliability.
Different parameters showed varying degrees of correlation for different
scenarios, runs, and speeds. However, one clear trend was the high tendency for
detection differences between the simulated model and the real model, with the simulated
model outnumbering the real model. Based on this, the developed method
can be used to some extent to relay information about whether the radar model is
trustworthy. However, its effectiveness heavily depends on the specific area of usage.
In conclusion, the method effectively visualizes the detections, indicating where they
occur and how they differ in number. While the model may not be suitable for final
certification, it can be a valuable tool for statistical approximation of the radar
model’s performance
Nya applikationer för kooperativ perception
No full textAs connected and autonomous vehicles become more prevalent, cooperative perception is increasingly vital for ensuring traffic safety and efficiency in urban environments. This project investigates how cooperative perception can enhance intelligent transportation systems by reducing collision risks and improving traffic efficiency through real-time data exchange between vehicles and infrastructure.
The system is developed in a simulation framework based on SUMO called ms-van3t. Which also integrates decentralized communication, collision avoidance, and route planning. We introduce a new concept called ITS-fairy, which uses a Server Local Dynamic Map to support real-time data sharing and hazard detection. In addition, a safety mode based on Time-to-Collision and Space-to-Collision estimations provides proactive warnings to prevent crashes. Additionally, a centralized planner is introduced to reduce congestion by assigning routes based on global traffic conditions.
The implementations were evaluated through simulations of urban traffic scenarios such as intersections and roundabouts. The results demonstrate ensured safety in urban scenarios and more coordinated traffic behavior under high-density conditions. We conclude that the developed platform offers a robust foundation for continued research and development in cooperative vehicular systems
Advancing Electrophoretic Deposition for Multifunctional Structural Batteries
No full textThis thesis presents a comprehensive study on the application of electrophoretic deposition (EPD) for the fabrication of multifunctional cathode electrodes in structural batteries. Structural batteries are emerging as a promising class of energy systems capable of combining mechanical and electrochemical functions in a single architecture. Such dual-purpose capability is of growing interest in sectors where lightweight design and space efficiency are critical, such as aerospace, automotive, and portable electronics. In this context, EPD was selected as the core fabrication method due to its advantages in process scalability, material versatility, and the ability to directly deposit active materials onto conductive structural substrates such as carbon fibres (CF).
The study was divided into four major experimental paths: (1) enhancement of deposition quality through magnetic field assistance during the EPD process; (2) improvement of
manufacturing efficiency via a redesigned high-throughput electrode holder; (3) extension of EPD application to electromagnetic interference (EMI) shielding by depositing Fe3O4-based composites; and (4) evaluation of electrochemical performance through controlled variation of reduced graphene oxide (rGO) content in LFP-based cathode formulations. While magnetic fields were not the main focus of this work, their selective application during deposition and drying stages proved effective in improving coating uniformity and reducing agglomeration under certain composition conditions.
To explore multifunctionality beyond energy storage, Fe3O4 was introduced as a magnetic filler material in the EPD suspension to fabricate EMI shielding electrodes. Although
EMI shielding effectiveness was not evaluated due to time limitations, the successful deposition of Fe3O4-based coatings on CF substrates supports the feasibility of EPD for
future dual-functional applications. In parallel, the study evaluated electrochemical performance using half-cell pouch assemblies, tested through open circuit voltage (OCV),
cyclic voltammetry (CV), electrochemical impedance spectroscopy (EIS), and galvanostatic cycling. The results indicated that moderate additions of rGO improve internal
resistance and electrode kinetics, while excessive carbon additives adversely affect coating
consistency and dispersion.
Overall, this work demonstrates the adaptability and versatility of EPD in fabricating structural battery electrodes, while proposing pathways for integrating electromagnetic
and electrochemical functionality. The results provide practical insights for future research and development of multifunctional power systems, laying a solid foundation for scalable, lightweight energy solutions in advanced engineering applications
Uncertainties and Design Margins A robust design approach for jet engine component design
No full textAerospace industry is driven by the need to develop new concepts and methods to
handle the constraints of weight and performance efficiency, reliability, regulatory
safety compliance, and cost-effectiveness. In parallel to these demands, engineers
have to manage increasing design complexity by using Multi Disciplinary models and
accelerate the product development cycles to be able to fulfil the market demands.
To achieve high performing, robust and sustainable product design, a more informed
management of design margins is required that minimizes or reduces the uncertainty.
First part of study has employed a qualitative methodology where interviews have
been conducted with experienced engineers. The interviews have focused the understanding
and practical experience on use of design margins, uncertainty quantification.
Furthermore, feasibility of adopting probabilistic tools in their day-to-day
engineering workflows have been asked. The data gathered through the interview
study has been analyzed and presented in AIM diagrams to establish the case for
a computational study. It has been observed that the decisions on design margins
are implicit and not in detail recorded, but are following previous design practices
connected to the area. These challenges are addressed by in the second part of the
study, where a generalized probabilistic framework are used into the existing designanalysis
environment at GKN. In this work it is realized by using the workbench of
ANSYS OptiSLang, which contains a workflow including sensitivity analysis to be
able to assess parameters influence to the response, a deterministic design optimization
to find the optimal combination of values for parameters. This can be finalized
by robustness and reliability analysis to ensure the considered design satisfies the
user defined robustness criteria expressed in terms of six sigma. The scope of the
thesis further extents to show that robustness analysis can also be studied by integration
of the mathematical framework of Probabilistic VMEA (variation mode
and effect analysis) into the workflow of ANSYS OptiSLang. This integration is a
challenge as well as an opportunity to make a more efficient algorithm for robustness
analysis. In this study, a steel hook and a simplified steel lug (used in aerospace
engines) have been used to illustrate the methodology with comparative results as
well as showing the opportunities by using Probabilistic VMEA