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.

Senast publicerade

  • Bild (thumbnail)
  • Bild (thumbnail)
    Intelligent Sound Diagnostics - Acoustic Fault Detection in Truck Steering Systems Using Machine Learning and Deep Learning
    (2026) Krishnamoorthy, Kiruthiga
    This thesis investigates acoustic fault diagnosis for heavy duty truck steering components, focusing on two safety critical parts, the Power Steering Pump (PSP) and the Steering Gear, using a real world dataset of 326 recordings collected under operational conditions. The problem is formulated as a seven class classification task covering four sound types under normal and faulty conditions. Classical machine learning models trained on hand crafted acoustic features are compared against a convolutional neural network trained from scratch and Pretrained Audio Neural Network (PANN) models used as frozen feature extractors, to identify which modelling approach is most effective for this task. Seven individual acoustic descriptors, MFCC, log-Mel spectrogram statistics, sub-band energy ratios, spectral shape descriptors, time-domain statistics, amplitude envelope kurtosis, and discrete wavelet transform coefficients, are also evaluated in isolation to determine which is most discriminative for fault detection. All models are evaluated using StratifiedGroupKFold cross validation with file level grouping to prevent leakage between overlapping audio segments, and performance is reported using the Matthews Correlation Coefficient (MCC) and macro F1, both robust to class imbalance. A linear Support Vector Machine trained on a 180-feature ANOVA selected set achieves the strongest overall performance, reaching a held out test MCC of 1.000, ahead of every deep learning approach evaluated, including a pretrained PANN backbone with over 80 million parameters. A compact CNN trained entirely from scratch matches the best pretrained PANN variant despite requiring no external pretraining, while the largest pretrained backbone performs weakest of all models compared. Among the individual descriptors, log-Mel spectrogram statistics and MFCCs prove the most discriminative, while amplitude envelope kurtosis alone carries almost no discriminative power. These findings suggest that carefully engineered acoustic features combined with a simple classical classifier can outperform learned and transferred deep representations on a dataset of this scale, offering a lightweight and interpretable alternative for industrial predictive maintenance.
  • Bild (thumbnail)
    Deep-EC2: A Non-Myopic Active Feature Acquisition Framework
    (2026) Sun, Haozhe
    Active Feature Acquisition (AFA) addresses the problem of sequentially deciding which features to observe in order to maximise predictive accuracy while minimising acquisition cost. Non-myopic methods, which account for how an acquisition changes the value of later ones, typically rely on reinforcement learning (RL), with its reward design, rollouts, and hyperparameter sweeps. This thesis proposes Deep-EC2, an AFA framework that replaces reinforcement learning with supervised regression onto analytical targets derived from the Equivalence Class Edge Cutting (EC2) criterion, the only AFA objective known to be adaptive submodular. A value network learns to rank candidate features by their expected EC2 gain, estimated with a masked k-nearest-neighbour estimator over the training set, and a third training phase adapts the predictor to the subsets that the learned policy acquires. At inference, acquisition requires only greedy forward passes, with no rollout or planning. Experiments on eleven benchmarks—three synthetic gated tasks, the CUBE-NM context-feature benchmark, and seven real-world tabular datasets—compare Deep- EC2 with seven published AFA methods, the RL method Opportunistic Learning (OL), and a random reference, under identical splits, three seeds, and seed-paired significance tests. On CUBE-NM, Deep-EC2 acquires a context feature first on every instance, significantly outperforms every baseline including OL, and reaches perfect accuracy after 7–8 acquisitions. On the synthetic task Syn3 it acquires the gate first on every instance and is on par with the strongest non-RL baselines. On both benchmarks a budget-by-budget analysis reveals an invest-then-exploit signature— an early deficit incurred by acquiring an individually uninformative context feature or gate, repaid by significant mid-budget gains—the hallmark of non-myopic acquisition. On the real-world tabular data Deep-EC2 remains competitive, and adapting the predictor to the policy improves all eleven benchmarks. Training a model takes 8–18 minutes on the synthetic benchmarks, faster than OL on each of them and about six to eighty times cheaper once OL’s hyperparameter sweep is counted. Overall, the thesis shows that non-myopic acquisition can be obtained from a supervised value network trained on EC2 targets, at myopic inference cost and without reinforcement learning.
  • Bild (thumbnail)
    AutomatingMaintenanceWorkOrder Pro cessing in Facility Management - Classification, Responsibility Routing, and Data Readiness in Maintenance Workflows
    (2026) Brown, Megan
    The Facility Management (FM) industry generates vast amounts of unstructured text data in the form of Maintenance Work Orders (MWOs). However, this data is frequently underutilised, often treated as an administrative burden rather than an analytical resource. Compared to hardware-dependent IoT (Internet-of-Things) and BMS (Building Management System) data, MWOs do not rely on modern building infrastructure and they can therefore provide valuable insights into building oper ations across diverse real-estate portfolios containing buildings of varying technical maturity. The Goal of this thesis is to investigate how MWOs can be leveraged to automate ad ministrative workflows and improve data quality within FM. Using a Design Science Research (DSR) approach in collaboration with Akademiska Hus, an exploratory data analysis established that traditional classification methods struggle with his torical noise, overlapping categories, and the under-specified nature of occupant text inputs. To address these bottlenecks, two artefacts, The Structurer and The De marcation Agent, were developed and evaluated. The results indicate that while full autonomy is restricted by non-deterministic model behavior and incomplete technical details in user texts, a Human-on-the Loop (HOTL) configuration provides a viable operational framework. Shifting the technician’s role from manual triage to validation lessens administrative burden without the risk of automated routing errors. Furthermore, the pipelines introduce a standardised classification logic that reduces the subjectivity and variance inher ent in manual logging, thereby improving data consistency for future maintenance analytics.
  • Bild (thumbnail)
    Climate Impact Assessment of Temporary Works in Urban Infrastructure - A Case Study of the West Link Project, Gothenburg
    (2026) Bähr, Olof
    Temporary works are essential for the construction of permanent infrastructure but are often excluded from climate assessments and procurement requirements. In complex urban infrastructure projects, such as underground rail construction, these temporary constructions may involve substantial quantities of materials, transport activities and machinery use, resulting in potentially significant greenhouse gas emissions. This thesis investigates the climate impact of temporary works in the Västlänken railway tunnel project in Gothenburg and evaluates how well Trafikverket’s climate calculation tool, Klimatkalkyl, represents these impacts. Four temporary constructions were selected for analysis: a diaphragm wall system, a crane foundation, a canal infill and an aqueduct. A life cycle assessment (LCA) was conducted using a three-scenario approach. Scenario A represents a standard Klimatkalkyl calculation based on generic default values. Scenario B adds construction activities that are structurally absent from standard calculations while maintaining the same emission factors. Scenario C applies project-specific Environmental Product Declarations (EPDs), transport distances and fuel data. The results show that the significance of omitted construction activities varies substantially between construction types. For the diaphragm wall system, the addition of previously excluded machinery activities increased total emissions by only 0.1%, whereas the aqueduct showed an increase of 137%. Applying project-specific data reduced calculated emissions by up to 45%, primarily due to lower-emission concrete mixes and the use of HVO100 in construction machinery. The findings indicate that current infrastructure climate calculations may underestimate the climate impact of machine intensive temporary works and that a gap exists between reported and actual construction phase emissions. While Klimatkalkyl can technically accommodate these activities, the current procurement framework does not systematically require their inclusion. The study concludes that this gap is structural rather than a question of data quality or compliance, and that temporary works require greater attention in climate reporting and procurement if long term climate-neutral infrastructure ambitions are to be supported.