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

AI i produktutveckling - Hur långt kan vi gå?
(2026) Dahl, Elin; Egert, Linnéa; Frölund, Emma; Holm, Ludvig; Zäll, Filippa
Artificiell intelligens (AI) blir allt vanligare inom produktutveckling. Syftet med detta projekt var att undersöka i vilken utsträckning AI kan bidra till ökad effektivitet och kvalitet, samt vilka begränsningar tekniken kan medföra när den integreras som en del av arbetsprocessen i produktutveckling. Under projektet har AI:s roll undersökts genom både manuella och AI-stödda arbetsmoment för att jämföra resultaten utifrån faktorer som tidsåtgång, relevans, kreativitet och behov av mänsklig korrigering efter validering. AI användes bland annat vid informationsinsamling, konkurrentanalys, behovsidentifikation, konceptgenerering och konceptutvärdering och analys av resultat. Detta undersöktes genom att arbeta med ett produktutvecklingsprojekt som utgår ifrån ett identifierat problem på Chalmers tekniska högskola. På högskolan förekommer ofta begränsade arbetsytor och instabila underlag, vilket skapar oro att dryck ska spillas över datorer och studiematerial. Utifrån detta problem utvecklades en fysisk prototyp i syfte att minska risken för spill av dryck i trånga studiemiljöer. Resultatet visar att AI kan vara värdeskapande för moment som kräver att större mängder information och data sammanställs, struktureras och analyseras. AI visades vara effektiv för framtagning av tabeller och för att föra underbyggda resonemang, vilket ger en tydlig överblick och ett användbart underlag för beslutstagande. Dock kräver AI:s resultat fortfarande mänsklig validering och bearbetning då resultatet ofta innehåller tekniska brister, feltolkningar av uppgiften eller bristande anpassning efter den specifika kontexten. Detta syntes tydligt i moment som konceptgenerering, CAD och detaljkonstruktion där den mänskliga erfarenheten och förståelsen var avgörande för utfallet. Den primära slutsatsen är att AI kan användas som ett värdefullt stöd i produktutvecklingsprocessen, men inte som självständig ersättning för mänsklig kompetens. AI kan bidra till effektivisering av uppgifter i vissa fall men slutliga bedömningar och beslut bör fattas av människor med relevant kompetens. Projektet visar även att utformningen av prompten som skickas in till AI är avgörande för hur resultatets utfall blir avseende relevans, kvalitet och användbarhet. Sammanfattningsvis kan AI på många sätt vara ett stöd i produktutvecklingsprocessen, men det krävs fortsatt mänskligt omdöme för att säkerställa att lösningarna som tas fram är realiserbara, korrekta och relevanta för problemet.
Early Classification of Silent Emergency Calls - Exploratory Audio-Based Detection of Pocket-Call-Like Acoustic Patterns
(2026) Granli Holmberg, Pontus; Reimertz, Kristoffer
Silent calls to the Swedish emergency number 112 place unnecessary strain on emergency-call operations and may delay responses to critical incidents. This thesis investigates whether accidental silent emergency calls can be identified using only caller-side audio. In particular, the work explores the classification of a manually defined subgroup of ”pocket calls”. Two audio feature representations were evaluated: handcrafted acoustic feature representations and pretrained self-supervised BEATs embeddings. Logistic regression and XGBoost classifiers were evaluated across multiple experimental settings, including delayed-speech scenarios, respiratory-audio analysis, and early detection using short audio intervals. The results showed that classification of the complete ”Silent 112” operator-assigned category resulted in poor and unstable performance, while significantly stronger performance was achieved for the ”pocket call” subgroup. Across all experiments, BEATs embeddings outperformed handcrafted features. The strongest performance was achieved using BEATs embeddings together with logistic regression on a 10- second audio window, achieving a recall of 0.92 with no false positives on the heldout test set. Additionally, reliable classification became increasingly feasible after approximately 6 to 8 seconds of available audio. Although the findings are preliminary and limited by the relatively small number of manually identified ”pocket calls”, the results demonstrate the potential of callerside acoustic analysis for supporting automated identification of accidental silent emergency calls. Keywords:
Development and Testing of Advanced Door Opening and Closing Function
Kalayaperumal Ravi, Senthil Kumaran; Campo, James
The increasing electrification and digitalization of heavy-duty vehicles have created new opportunities to improve both vehicle functionality and aerodynamic performance. Reducing aerodynamic drag is an important criterion identified for lowering energy consumption and increasing vehicle range, and the integration of flush exterior components, such as door handles, can contribute to reducing surface discontinuities. This thesis investigates the development of a power-operated side-door system for a heavy-duty truck, with a focus on integrating power release, soft-close functionality and a flush exterior handle interface while retaining the existing latch mechanism and mechanical release as a backup. A structured product development process was used to gauge the user needs and understand contemporary door entry systems to then develop and evaluate the prototype. The existing door system was integrated with an electric linear actuator in place of the door checker to enable a multi-functional solution. The actuator mounting geometry, required stroke, opening angle and operating forces were determined through kinematic and force calculations. Furthermore, a geared motor-driven mechanism was developed for power release and cinching. The multi-stage gear system was developed to provide the required output torque to the latch, and the drivetrain was verified using Romax. The housing, gears and opening interface were subsequently refined through several prototype iterations and functional testing. A preliminary communication and control system was also developed and tested using an Arduino-based setup to coordinate the different components and demonstrate the operating sequence. The final prototype combined the developed actuator, latch interface, cinching and powerrelease functions, exterior handle interface and supporting test fixture into a functional system. The iterative development and testing process identified several practical limitations related to packaging, structural strength, electrical interface and component integration, which were addressed through successive design modifications. The results demonstrate the feasibility of integrating an electrically operated door system around an existing heavy-duty vehicle door, while the developed prototype provides a technical foundation for further refinement and integration toward a production-ready power-operated door system for heavy-duty vehicles.
Design Optimization and Dynamic Analysis of engine mounts and flanges of a marine drivetrain
(2026) Niranjan, Brijesh
In reality, marine vessels undergo various dynamic loads while breezing through the waters and crashes into the waves. This leads to g-forces experienced by the marine vessel. A marine vessel consists of the main engine, electric machine, and the gearbox coupled by flanges and their respective brackets mounted to the vessel by rubberized mounts. Due to varying loads, the brackets and the mounts experience forces on them and, in return, produce reaction forces. In addition to this, the weights of each component contribute to the moment of the system at the flanges. An analytical method was developed in [1] to estimate these forces on the mounts and the moments on the flanges. This thesis focuses on optimizing the design to minimize these loads experienced and to conduct a dynamic analysis of the whole system.
Kompakt mikrovågssystem för detektering av intrakraniella blödningar
(2026) Agfors, Lovisa; Bergman, Naëmi; Ek, Andreas; Ingels, Virve; Murén, Izabella; Myllykangas, Lina
Stroke is one of the leading causes of death in Sweden and worldwide. Despite the fact that the time between symptom onset, diagnosis and treatment has a major impact on a patient’s chances of survival and recovery, there are currently no effective and reliable methods for prehospital stroke diagnosis. Microwave imaging has been proposed as a potential solution for detecting intracranial hemorrhages, such as stroke, already in the ambulance. However, these measurement systems are limited by factors such as high cost and bulky design. The aim of this project was to further develop and evaluate a compact and cost-effective measurement system designed to achieve performance comparable to that of a Vector Network Analyzer (VNA). To enable the measurement of weaker signals, the system was modified to reduce noise levels and crosstalk between system components. Different configurations using different components were tested and evaluated with respect to signal-to-noise ratio, isolation, and the ability to detect weaker signals. To evaluate the performance of the system and to develop a simple machine learning algorithm, a phantom model of a human head both with and without hemorrhage was constructed. The measurement results showed that the isolation of the measurement system had increased from 30 dB to 100 dB, and that its ability to detect weaker signals had improved. The measurement system and the VNA produced comparable measurement results for phantoms with and without hemorrhage. However, the measurement system still exhibited higher noise levels and lower precision compared to the VNA, indicating that further development is necessary. Although it was difficult to visually distinguish between measurement results from phantoms with and without hemorrhage, the classification algorithm achieved a high accuracy (81%) and an AUC value of 0.963.