Chalmers Open Digital Repository

Välkommen till Chalmers öppna digitala arkiv!

Här hittar du:

Studentarbeten utgivna på lärosätet, såväl kandidatarbeten som examensarbeten på grund- och masternivå
Digitala specialsamlingar, som t ex Chalmers modellkammare
Utvalda projektrapporter

Forskningspublikationer, rapporter och avhandlingar hittar du i research.chalmers.se

Enheter i Chalmers ODR

Välj en enhet för att se alla samlingar.

Visar 1 - 2 av 2

Senast publicerade

Safe AI Compliance in CI/CD for Software-Defined Vehicles - Researching Conformity to ISO/PAS 8800:2024
(2026) Daneshvar-Minabi, Atosa; Heijkenskjöld, Ann
As software-defined vehicles (SDVs) are incorporating artificial intelligence (AI), ensuring AI safety presents significant challenges for development teams. This thesis addresses four research questions spanning the current application of ISO/PAS 8800:2024 Road vehicles — Safety and artificial intelligence in SDV development, the compliance challenges teams face, where in the Continuous Integration/Continuous Delivery (CI/CD) challenges arise, and how to mitigate them. A two-phase qualitative interview study was conducted with industry professionals, including do main experts and researchers. Through semi-structured interviews, recommendable actions were produced and evaluated in the form of artefacts and a set of guidelines. The results reveal that while some individual practitioners are aware of ISO/PAS 8800:2024, formalised organisational conformance frameworks are largely absent, and a variety of different conformance issues surface throughout the entire CI/CD process rather than at any single stage. Artefacts were developed for supporting conformance to ISO/PAS 8800:2024 within existing development practices. The first artefact guides teams on where human oversight is required, where AI could be helpful, and where AI automation is feasible across the AI safety lifecycle, including the trade-offs and acceptance of AI automation at each stage. The second artefact shows which selected clauses of the ISO/PAS 8800:2024 are most prominent at each phase of the MLOps lifecycle. Integrating ISO/PAS 8800:2024 into CI/CD requires conformance to be treated as an inherent part of development rather than an after-the-fact activity. Insights from this study could be valuable for organisations integrating ISO/PAS 8800:2024 into their development workflows.
Improving Inbound Logistics through Standardised Packaging
(2026) Kristensen, Isak; Rudresh, Eshwar
This thesis focuses on how standardised packaging affects inbound logistics, using the example of Sansera Sweden AB in Trollhättan, a manufacturer of precision engine components for the automotive industry. Sansera sources components from four suppliers, three of which use the standardised V-EMB containers (standard-shaped, stackable, returnable modules), while one supplier delivers parts in non-standard, non-stackable, and disposable cardboard boxes. Based on three interviews with Sansera’s logistics coordinator, observations of Sansera’s inbound logistics processes, and 2025 transportation data, the analysis identifies the non-standard supplier’s packaging as a common cause of inefficiency across several performance measures. These measures include a trailer load utilisation of approximately 20 percent versus 95 percent for suppliers using standardised packaging, over four times the C02 footprint per part compared to the standardised flow, increased warehouse capacity pressures, and recurring production disturbances amplified by the company’s sole-source reliance on this supplier. The thesis concludes that packaging must be considered part of a company’s system-level inbound logistics decisions and proposes that the non-conforming supplier could be offered standard reusable packaging delivered through the existing reverse empty-packaging flow from another supplier.
Energy-Aware TinyML for Ambient-Powered, Hardware-Constrained IoT Nodes
(2026) Krishnavilasom Gopalakrishnan, Anakha
Resource-constrained Internet of Things (IoT) devices are increasingly expected to perform intelligent sensing under strict limitations in memory, computation, and energy. For image-based applications, transmitting raw images can consume substantially more communication energy and time than transmitting a compact latent representation. This thesis investigates lightweight convolutional autoencoders for on-device image compression on microcontroller-class hardware. A design space of 70 convolutional autoencoder models was evaluated on the MNIST dataset. The models differed in encoder filter configuration, bottleneck dimension, and compression ratio. Reconstruction quality and task utility were measured by normalized mean squared error (NMSE) and downstream classifier accuracy. The baseline models were further optimized using post-training quantization (PTQ), quantization-aware training (QAT), and a two-stage magnitude-based pruning method. Different optimization orders were compared with respect to model size, reconstruction quality, classifier accuracy, and estimated energy. An analytical energy model was used to estimate microcontroller active energy associated with encoder computation and payload transfer under explicitly stated assumptions. Pareto-front analysis identified non-dominated candidate models offering different tradeoffs between reconstruction quality and estimated energy. A selected INT8 encoder was deployed on an Ambiq Apollo3 microcontroller using TensorFlow Lite for Microcontrollers and the NeuralSPOT software stack. Hardware experiments confirmed successful encoder inference and provided measured inference-time results. The results show that bottleneck size and the choice of optimization method strongly influence the trade-off between reconstruction quality, payload size, and deployment cost. Quantization reduced numerical precision and memory footprint, while unstructured pruning introduced sparsity but did not necessarily reduce execution time on dense microcontroller kernels. The study therefore carefully distinguishes analytical energy estimates from measured hardware results. This work provides a reproducible framework for comparing learned compression models on resource-constrained embedded devices. The evaluation is limited to MNIST data, analytical communication-energy assumptions, host-side serial data transfer, and a single microcontroller platform. A complete autonomous energy-harvesting and wireless IoT
Oil Quality Sensor (OQS) Sensor performance testing and modelling
(2026) Talha, Mohammed
Engine oil condition in Volvo Group heavy-duty field trucks is currently assessed through fixed interval oil changes supplemented by laboratory sample analysis, a process that takes up to four weeks and cannot detect real time contamination events such as water contamination or fuel dilution before damage occurs. This thesis investigates whether the TE Connectivity FPS2800 Oil Property Sensor measuring viscosity, density, dielectric constant, and electrical resistivity simultaneously is capable of replacing this process with real-time oil monitoring for Volvo VDS-5 engine oil. The work was carried out in four phases starting with the sensor’s temperature dependent properties which were validated on a oil test rig resulting in three of the four proposed models (viscosity, density, dielectric constant) held in line for VDS-5 oil, while resistivity required a corrected Arrhenius cubic model to replace the originally proposed linear one. Fuel dilution models were calibrated at 3%, 6% and 9% and water contamination at 1% and 2% concentrations. further on to second phase where the validated models were applied to three months of CAN-bus data from a Volvo field test vehicle by building a pipeline on Volvo’s Data Science Lab. Moving to third phase where the resulting oil condition estimates were validated against Volvo’s laboratory reference data which showed agreement for soot, water contamination and oxidation. Finally the fourth phase where the pipeline was deployed end to end with a Power BI dashboard for research and development team. The results show that the FPS2800, calibrated specifically for VDS-5 oil, can deliver oil condition estimates comparable in accuracy to laboratory analysis for the most critical parameters, furthermore, estimated prototype cost was estimated and business case was made for the sensor. Remaining gaps like oxidation and soot rig calibration are identified as future work ahead of production deployment.
Strategic Alignment in Servitization Transformation: A Retrospective Case Study
(2026) Eriksson, David; Friman, Erik; Gustafsson, Johan; Liljeström, Adrian; Marthinsson, Oskar; Österling Junkargård, Måns-Perra
Problem Despite the extensive body of servitization research, there remains a limited understanding of how transformation processes unfold over time within firms. Prior studies have primarily emphasised the importance of deliberate strategic efforts for enabling successful servitization. More recent research, however, increasingly suggests that servitization also unfolds through emergent organisational adaptation over time. Nevertheless, how deliberate and emergent strategy interact in practice, and how this interplay shapes strategic alignment throughout the transformation process, remains insufficiently understood in the literature. Aim The aim of this study is to explore how deliberate and emergent strategy interact over time within a servitization transformation, and how associated alignment processes shape the transformation trajectory. In doing so, the study seeks to address the identified research gap regarding processual dynamics of servitization transformation. Theoretical framework This study draws on servitization literature and a strategy process perspective, where transformation is understood as the interplay between deliberate and emergent strategies and where strategic alignment constitutes a central mechanism. It further applies dynamic capabilities theory to explain how organisations adapt over time through sensing, seizing, and reconfiguring activities. These perspectives are integrated into an adapted theoretical framework illustrating how the strategy process unfolds in a servitization transformation. Method The study adopts a qualitative and exploratory research design, conducted as a single-case study of a division within a manufacturing firm. Data was primarily collected through semi structured interviews with respondents from multiple organisational levels and functions, complemented by secondary data. The findings were first analysed using an adjusted thematic analysis and subsequently interpreted through the adapted theoretical framework, following an abductive process where theory and empirical data are iteratively linked. Results and implications The study shows that the servitization transformation unfolded through an ongoing interplay between deliberate and emergent strategy, where the relative influence of each varied depending on contextual conditions. Emergent developments, particularly driven by customer demand, played a central role in shaping the realised strategy throughout the transformation. At the same time, deliberate strategy remained important for enabling coordinated reconfiguring and sustaining strategic alignment between intended and realised strategy over time. Rather than developing through stable implementation processes, alignment evolves unevenly through periods of disruption, partial alignment and broader realignment. The study contributes to existing servitization literature by providing a more process-oriented understanding of how deliberate and emergent developments interact over time and shape strategic alignment throughout the transformation process. The findings further suggest that a high degree of customer involvement may intensify the role of emergent strategy within servitization contexts. From a managerial perspective, the study highlights the importance of enabling continuous organisational adaptation and developing reconfiguring capabilities.