Chalmers Open Digital Repository
Not a member yet
26247 research outputs found
Sort by
Generative Models for Context Dependent Urban Planning
No full textBuilding footprints represent the total area of coverage of a physical building. Footprints
can be used to give an overview of a planned developed area in the early stages
of urban planning. This thesis investigates the possibility of training a generative
AI model to generate building footprints in a designated area based on surrounding,
already existing building footprints. Such a generative model would be a useful tool
for architects in the early stages of urban planning, as it would allow the rapid generation
of footprint suggestions in an area designated for development. Two image
inpainting networks trained for general image reconstruction were fine-tuned using
a dataset of building footprints to improve on the task of footprint generation. One
of the networks was also modified and trained to be able to accept a desired density
of footprints in the generated area as an additional input. Two different masking
algorithms were used during training and evaluation: A simple square approach and
a more sophisticated algorithm that finds and masks city blocks. FID and LPIPS
were used to evaluate and compare the trained models. It was shown that image
inpainting networks form a good basis for context dependent footprint generation
and that fine-tuning improves performance on this task. Furthermore, it was demonstrated
that an image inpainting network can be modified to accept and adhere to
density requirements providing a proof-of-concept for other types of user guidance
Water Tree Mapping in Submarine High Voltage Cables Using Finite Element Method
No full textAbstract
Water treeing is a degradation phenomenon that occurs in submarine power cables. It is one of the main causes of failures of dynamic cables attached to mobile floating platforms. The rough sea environment causes mechanical stresses in the insulation of dynamic cables provoking appearance of microscopic cracks and water intrusion. The combination of water and enhanced electric fields can cause electrical, chemical, and mechanical reactions to occur which intensify water tree growth in cable insulation. This thesis aims to contribute to the development of a numerical model for simulating water treeing in high voltage insulating materials. The scope of the thesis is limited to the electrical aspects of the water treeing. COMSOL Multiphysics is used to model the conditions for experimental water tree testing according to ASTM D6097 standard. The model considers the current flow in the polymeric insulation due to the applied electric stress and so-called state variable is used to map healthy and damaged regions in the insulation material. The latter is identified as a domain where the electric field exceeds the threshold corresponding to initiation of defects in the polymer due to appearing internal electrostatic forces. Validation is done by comparing the simulation to the standard and experiments reported in the literature. The results show that the numerical model follows previous observations and may be used as a base for further development of the tools for predicting insulation lifetime
Design of smart orthosis for rehabilitation of Achilles tendon ruptures
No full textAbstract
This thesis presents a compact system for measuring forces under the foot, designed for rehabilitation after an Achilles tendon rupture. The aim is that this system could provide patients and care providers with critical data about the recovery process, for a more personalized and effective treatment. The core component of this system is a newly developed flexible insole that senses forces under the foot. The force is measured in three dimensions (i.e., normal
and shear forces) using magnetic-based sensors, placed in a grid of 73 nodes. The large area covered by the sensor, and the flexibility, are improvements over previous magnetic-based force measurement systems. The insole and additional support electronics were mounted on a standard ankle orthosis (also known as Walker). In addition, two IMUs were used to estimate the orientation of the insole. Software was also developed to process and visualize the data. The measurements from the insole are sent to a signal processing chain to calculate relevant biomechanics parameters such as the center of pressure and joint torques. The signal processing chain was implemented within ROS2, together with micro-ros for the low-level communication with hardware. ROS2 is also used for visualization purposes. The results are promising, showing that magnetic-based sensors are feasible for measuring 3D forces under the foot. The sensors display a nearly linear response to vertical pressure, although there is considerable hysteresis that introduces errors in the measurements. Future work to improve calibration, verify reliability, and improve ease of use is needed before the system can be used in a clinical setting
End-of-Life Management for Digital Battery Passports in Electric Vehicles
No full textThe use of batteries increases every year, and it is more important now than ever to transition into a sustainable way of handling them in the right way. This requires traceability along the whole battery value chain, including information from mining raw materials until the battery is recycled. The EU has implemented the battery regulation, stating that a digital battery passport (DBP) will be required from February 2027. The DBP enables traceability of the battery and facilitates the recycling process. However, the DBP system is not yet finalized, and research shows that there is a gap in guidelines for how it should be managed during its end-of-life (EoL). Therefore, this project aims to determine how to manage DBP for Li-ion electric vehicle batteries (EVB) in the EoL phase to enhance traceability, enable a circular economy and enforce a sustainable transition. This project investigates the complete battery value chain to understand the EoL. Through a literature review and interviews, this thesis dives deeper into the management of batteries, appropriate solutions of ending its corresponding DBP and loopholes that may occur. The results show that the responsible economic operator (REO) creating the DBP is the owner and the responsible actor during its whole existence and is the only one that has the right to end it. This means further that independent operators never will have REO responsibilities, on the other hand, they are obliged to update DBP when needed for the battery. A DBP can only end after its product has been recycled and when this occurs, the recycling station will automatically inform the REO to end the DBP. A two-step verification process controls each individual product and informs its status to the corresponding REO and reduces the risk that information for some products would not be documented. For the recyclers who receive information about all products, this thesis presents a solution for module identification (MID), enabling modules to carry information, since the DBP only follows the complete battery pack. With the two-step verification process, the proposal with MID and keeping the responsibility by the REO, the loopholes regarding lack of information on products within the system can be significantly reduced. In conclusion, this thesis provides guidance on how to manage the DBP in the EoL phase based on currently available information and decisions on the topic
Next Generation Steering Device - Designing for Emergency Situation
No full textAs autonomous vehicles progress toward full automation, new challenges arise in ensuring safety during emergency scenarios where human intervention may still be
necessary. This thesis, conducted in collaboration with Autoliv, investigates how Level 4 robotaxis can be designed to allow first responders to manually reposition
stalled vehicles in critical situations. The project focuses on developing a fallback steering solution that addresses the absence of traditional controls in future vehicle
interiors. Using a human-centered design process, including literature reviews, in-depth interviews with automotive experts and users, and iterative prototyping, the
project identifies key user needs such as intuitive operation, mechanical reliability, and secure access.
The result is HALOGRIP, a visible, analog steering device embedded within the dashboard of the robotaxi. Activated via a two-step ID verification process, the device uses a tilt-based mechanism for speed control and traditional rotation for steering. The system enables low-speed maneuvers without requiring training and supports quick response in space-constrained environments. Physical buttons and a HUD provide clear operational feedback, while mechanical locking ensures failsafe deployment. HALOGRIP offers a pragmatic and user-friendly solution that empowers emergency personnel without compromising the autonomy of the vehicle. Future work includes validating usability in real-world scenarios and refining integration across different vehicle platforms
Analysis on the determinants of EV purchase intention in Sweden
No full textThis study uses a discrete choice experiment embedded in a survey to explore the
determinants of Swedish consumers’ choice between electric vehicles (EVs) and
internal combustion engine (ICE) vehicles. A total of 373 respondents, resulting in
7,266 valid choice observations, were collected and analyzed using binary logistic
regression models. Model specifications include both vehicle-specific attributes (e.g.,
price, range, maintenance costs, charging time, charging convenience, and emissions)
and demographic characteristics (e.g., age, gender, education, income, and family
structure).
The results show that economic and infrastructure considerations dominate consumers’
decision-making process. Specifically, vehicle price, maintenance costs, and the
availability of home charging infrastructure are significant attributes of EV adoption.
The existence of a home charger is a particularly important driver, increasing the
probability of choosing an EV by nearly 60 percentage points on average. In contrast,
attributes such as range and charging time, while in the direction of theoretical
expectations, are not statistically significant in the current sample. The probability
analysis also highlights that in the absence of home charging facilities, the impact of
price cuts is relatively limited, suggesting that policymakers should increase investment
in EV charging facilities.
The study provides practical insights for policymakers aiming to accelerate the adoption
of EVs in Sweden. In addition to targeted financial incentives, efforts should focus on
improving private and public charging infrastructure. The findings also contribute to a
broader understanding of how practical and infrastructure factors influence low-carbon
transport choices in European markets
Automated Validation of Test Cases Using Generative AI: Development of program for generating test cases utilizing generative AI for system requirements in automotive industry
No full textThis project was provided by Volvo and was intended to help automate their current system of creating validation test cases using generative artificial intelligence (AI). Automation of this system reduces development time and costs while enhancing efficiency. This project covered theories regarding AI, Large Language Model (LLM), prompt engineering, test cases and system requirements. Following an evaluation of available open-source LLMs, the model QWQ-32B was selected. By applying prompt engineering techniques, the model was able to generate not only validation test cases but also executable test code. The program was able to generate adequate results according to Volvo staff. However, the results needs some adjustments in order to be viable for a variety of test cases. Potential improvements, such as the integration of Retrieval-Augmented Generation (RAG), are discussed in this report as future directions to address current limitations. As a result of this project, Volvo has been provided with a solid foundation for automating generation of validation test cases and executable test code
Energy efficiency evaluation of the secondary heating system at the Södra Cell Värö pulp mill
No full textExploring Artificial Intelligence use in Services Procurement A case study at Volvo Group
No full textProcurement has evolved into a strategic function and emerging technologies, such as
artificial intelligence (AI) are digitizing supply chains, constituting a major digital shift in all
industries. Despite its potential, artificial intelligence is still both underutilized and under
researched within procurement. While existing research focuses on general applications of
AI in procurement, this thesis addressed the area of services procurement. By identifying
pain points across an established services procurement process and evaluating how AI can
address these, it aimed to explore how AI could be leveraged to enhance purchasing
efficiency. Additionally, it explored how change management and technology acceptance
strategies can support organizations in AI adoption.
The thesis has been conducted as a case study together with the Services Purchasing
organization at the global transport- and infrastructure provider Volvo Group. Taking an
explorative approach, unstructured interviews, primarily with line managers, and semi-
structured interviews with buyers were conducted. Following a thematic analysis of the
interviews, empirical findings were analyzed using theoretical frameworks developed
through an extensive literature review. The theoretical frameworks covered AI.
The empirical findings identified pain points across different stages of the services
procurement process, revealing both operational, organizational, and strategic challenges
affecting the efficiency of the procurement organization. Applying the theoretical
frameworks, the analysis showed how AI can support buyers and address several of the key
pain points in the services procurement process. Leveraging AIs main capabilities of
automation and smartness, the technology shows strong potential in e.g., automating
request-for-quotation creation and contract management, enhancing decision making, and
providing support in negotiations. Finally, trustworthiness, quality of output, job relevance,
ethics, and confidentiality were identified as requirements for the further adoption of AI
tools. Recommended change management strategies for successful AI adoption included
among others having local AI champions, sharing success stories, and developing a clear
vision and strategy