Chalmers Open Digital Repository
Not a member yet
26247 research outputs found
Sort by
Exploring AI Usage in Management Consulting Leveraging AI for Potential Benefits at the Intersection of Business and Technology
No full textAs generative AI reshapes industries, management consulting faces a transformation. This thesis explores how consulting firms can leverage AI strategically to enhance value creation both internally and externally. Through a qualitative case study of a small and relatively newly established Swedish management consulting firm, supplemented by insights from additional consulting firms and client perspectives, this study explored the changing consultant role. In doing so, the thesis also examines what competencies and capabilities will be necessary to remain competitive and relevant in the future.
The findings show that AI is not replacing human consultants, rather redefining their roles. With AI as an efficiency enabler, suitable for repetitive tasks, consultants can shift their focus to more value-enhancing parts of the project. The human aspects of consulting, referred to as soft skills and including skills such as communication and trust building, are aspects that AI is not able to fill. The study introduces the Human-AI Value Matrix, a framework for mapping how firms can position themselves by balancing AI integration with the irreplaceable aspects of human insights. The study further reinforces the concept of “Hybrid Consultants”, who combines AI literacy with domain expertise and emotional intelligence.
Ultimately, the thesis argues that competitive advantage in the AI era comes not from technical capabilities alone, but from consulting firm’s ability to integrate the advantages that comes with AI into a wider context. In order to truly gain value from AI, it is necessary to have someone who can translate the insights and apply them to the unique case of every client. This provides practical implications for consulting firms seeking to stay relevant and future-proof their offering
Hur företagsförvärv påverkar innovation i forskning- och utvecklingsintensiva branscher En kvantitativ studie om företagsförvärv och dess påverkan på forskning- och utvecklingsintensiteten inom svenska medicin och läkemedelsbolag
No full textDenna studie undersöker hur företagsförvärv påverkar innovation inom FoU-
intensiva branscher, med särskilt fokus på svenska företag inom medicinteknik- och
läkemedelssektorn. Studien använder en kvantitativ metod för att analysera hur FoU-
tillväxt och FoU-intensitet utvecklas före och efter att ett företag förvärvats, samt hur
dessa förändringar varierar beroende på typen av förvärvare och bransch. Paneldata
analyseras med hjälp av två olika random effects-regressionsmodeller, baserat på
data från svenska målbolag som förvärvats mellan 2000 och 2020. Resultaten visar
statistiskt signifikant att företagsförvärv har en negativ effekt på FoU-tillväxten i
målbolaget. FoU-intensiteten tenderar också att minska, men denna förändring är
inte statistiskt signifikant. Studien finner även att förvärv av industriella aktörer är
förknippade med större nedgång i FoU-tillväxt än förvärv av riskkapitalbolag. Det
finns också märkbara skillnader mellan branscherna i samband med FoU-intensitet.
Läkemedelsbolag uppvisar betydligt högre FoU-intensitet än medicinteknikföretag,
vilket bekräftar tidigare forskning om skillnader i utvecklingscykler och regulatoriska
krav. Dessutom visar studien att ett positivt börsklimat, mätt genom OMXS30-
indexet, har en svagt positiv effekt på FoU-investeringar. Dessa resultat ger insikter
i hur förändringar i ägarstruktur påverkar innovationskapaciteten och utmanar den
vanliga uppfattningen om företagsförvärv
Assessing the crash avoidance potential of cut-in crashes
No full textRoad traffic crashes remain a major safety concern. To reduce their occurrence, automated driving functions (ADFs) are being developed—but these systems must be thoroughly validated. Virtual simulations have been used to assess their performance, often comparing results between different systems without establishing a benchmark for what is theoretically achievable. This study focuses on cut-in crashes, a relatively common and challenging scenario for automated driving systems. It aims to assess the proportion of such crashes that could theoretically be avoided through braking alone and to compare this benchmark with the performance of two reference driver models and an ADF. This was achieved by developing an idealized model that reacts earlier and brakes harder than realistically possible, ensuring that no other model should be capable of outperforming it. This ideal model as well as the reference driver models and the ADF were then applied in virtual counterfactual simulations to estimate the proportion of crashes they could avoid. The cut-in scenarios simulated were categorized as either frontal or non-frontal cut-ins. The study was able to establish an upper limit for the non-frontal cut-in crashes but not for the frontal ones, as the ideal model avoided all frontal collisions. The two reference driver models avoided 38.5% and 81.8% of the frontal crashes, respectively, illustrating a large discrepancy. It remains unclear whether this reflects differences in the modeled driver’s behavior or limitations in how well the models represent real human drivers
Autoformalization for Agda via Fine-tuning Large Language Models
No full textAutoformalization, translating informal mathematical statements into fully formal machine-checkable code, has recently attracted interest thanks to advances in large language models (LLMs). However, progress has been hindered by the lack of highquality parallel corpora pairing natural language with formal code, especially for less-popular systems such as Agda. We address this gap by introducing SMAD (Synthetic Multilanguage Autoformalization Dataset), a 400K 4-to-3 parallel corpus
covering four formal languages (Dedukti, Agda, Coq, Lean) and three natural languages (English, French, Swedish), generated via the Informath project. We finetune the open-source Qwen2.5-7B-Instruct model on SMAD and achieve substantial gains: BLEU-4 improves from 32.90 to 76.16, and Agda syntax error rate falls from 98.43 % to under 8 %. We further explore joint training across multiple formal and natural languages, demonstrating that multilingual and multi-formal regimes yield notable improvements in low-resource Agda settings. Our work establishes the first LLM-based Agda autoformalization system and provides systematic insights into model scaling, multilinguality, and dataset construction for future research
Rethinking office space as a place to age; converting an office building into an assisted living facility for people with dementia using a case based approach
No full textWith rising numbers of people with dementia and
increasing office vacancies, this thesis explores how
existing office buildings can be repurposed to meet
the needs of Assisted living facilities for people with
dementia. The study addresses environmental and
social aspects through the example of Högsbo, an
area dominated by offices and warehouse buildings,
currently planned for transformation into a mixed-use
residential district.
The aim was to identify strategies for converting office
spaces into high-quality living environments for people
with dementia. Special attention was given to the
specific needs of residents, such as orientation and
comfort, while considering the building stock. These
included adapting the building stock and extensions
to enhance the building performance and user wellbeing.
Guided by the research question “Is it possible to
transform office buildings into high-quality living
environments for people with dementia?”, the thesis
applied an evidence-based design approach. Analysis
of reference projects and literature provided the
foundation for design strategies, further refined by the
Human-centered-design theory.
The results demonstrate the ecological and social
potential of adaptive reuse while highlighting key
challenges inherent in office typologies, such as narrow
floorplans and variation in floor heights. To address
these, spatial and structural strategies are proposed,
ensuring both functional and atmospheric qualities to
meet the demands of the residents with dementia.
By combining architectural adaptation with the
specific need of the target group, this work contributes
transferable insights for the ecological, economic,
and social sustainability of future office-to-housing
conversions
Dimensionering och utvärdering av energiförsörjningssystem för eldriven färja på sträckan Göteborg-Fredrikshamn
No full textAbstract
The electrification of the maritime sector plays a critical role in reducing emissions from transportation. This study evaluates the suitability of the energy systems, batteries and fuel cells for powering an electric ferry operating on the route Gothenburg - Frederikshavn. The evaluation is based on the key factors energy consumption, cost, and environmental impact. Additionally, the study examines the extent to which solar panels can contribute to the vessel’s energy supply. The study is based on the RoPax ferry Stena Jutlandica and is conducted through literature studies, expert interviews, and technical calculations based on real operational data. The study is limited to these three energy systems and does not include simulations or comparison with fossil-based systems. Two types of batteries, fuel cells and solar panels are compared and the result shows that LFP-batteries offers lower energy consumption, cost and environmental impact while Solid Oxide Fuel Cells with ammonia as fuel source have the advantage of energy capacity. Solar panels can provide a limited share compared to the other energy systems. Therefore they are not suitable as the primary propulsion source but can serve as a complementary system, contributing to the hotel load. In summary, the study shows that LFP-batteries, due to their energy consumption and cost is the most advantageous energy system for an electric ferry on the route Gothenburg–Frederikshavn. Furthermore, solar panels can contribute with energy to partially cover the hotel load. To reduce charging power, an energy system containing both batteries and fuel cells could be an interesting alternative for further investigation.
Sammandrag
Elektrifieringen av den marina sektorn är ett viktigt steg för att minska utsläppen från transportsektorn. Denna studie utvärderar vilket av energiförsörjningssystemen batterier och bränsleceller som är bäst lämpat för en eldriven färja på rutten Göteborg-Fredrikshamn, utifrån energiförbrukning, miljöpåverkan och kostnad. Det görs också en undersökning angående i vilken utsträckning solceller kan bidra med energi. Studien utgår från RoPax-färjan Stena Jutlandica och bygger på litteraturstudier, expertintervjuer och tekniska beräkningar baserat på verklig driftdata. Studien är avgränsad till att jämföra batterier och bränsleceller och omfattar inte simuleringar eller jämförelse med fossilbaserade system. Två typer av batterier, bränsleceller och solceller jämförs och resultatet visar att LFPbatterier erbjuder lägre energiförbrukning, kostnad och miljöpåverkan medan fastoxidbränsleceller med ammoniak som bränsle är fördelaktigt vad gäller energikapacitet. Solceller kan endast bidra med en liten mängd energi i förhållande till de andra energiförsörjningssystemen. Därför lämpar det sig inte som huvudsaklig framdriftskälla, utan endast som ett bidrag till hotellasten. Sammanfattningsvis visar studien att LFP-batterier med sin energiförbrukning och kostnad är det mest fördelaktiga energiförsörjningssystemet för en eldriven färja på sträckan Göteborg-Fredrikshamn. Dessutom kan solceller bidra med energi för att delvis täcka hotellasten. För att hålla nere på laddningseffekten kan ett energiförsörjningssystem bestående av en kombination av batterier och bränsleceller vara ett intressant alternativ för vidare undersökning
Utveckling och validering av en oljekoppling för turboaggregat
No full textThis bachelor’s thesis, conducted in collaboration with Volvo GTT, aims to develop an alternative design proposal for an oil connection to the turbocharger housing. The project was initiated to develop a component that is easier to assemble, has lower complexity, and reduced lead time. The existing solution is functional but suffers from high material consumption and complicated assembly, which motivates a new design.
The work follows a systematic product development methodology, beginning with a market analysis and a technical review of the current solution. Tools such as brainstorming, elimination matrices, Pugh matrices, and Kesselring matrices were then used to generate and screen concepts. The final choice fell on a simple, robust design that seals through metallic deformation during assembly.
CAD models and technical drawings were developed according to ISO standards and Volvo’s internal guidelines. The selected material was austenitic stainless steel due to its high hardness and temperature resistance. Alternative materials such as ferritic steel were also considered as potentially more cost-effective options.
A validation plan has been developed, including pressure, vibration, and salt tests, as well as thermal cycling. Finite Element Method (FEM) simulations were conducted to analyse contact pressure and plastic deformation in the seal. The results indicate that the load is absorbed as expected, but further simulations are recommended. For continued development, physical testing, FMEA updates, and economic and environmental analyses are proposed. In addition, a dialogue must be established with suppliers and Volvo GTO in Skövde. The project has delivered a proposal for an improved oil connection with the potential to streamline production and reduce costs while maintaining functionality and durability
Facterra
No full textA massive amount of waste is generated globally, and the circular economy offers a promising approach to reduce it by reusing resources and minimizing environmental impact. Raising public awareness is a key step toward supporting this transition. One potential method for raising awareness is through video games. This project aimed to design and develop Facterra, a 2D factory simulation game that is both fun and educational, with a focus on environmental impact and the circular economy. To evaluate whether the game achieved its goals, playtesting was conducted where players experienced the game and answered a combination of surveys and interview questions. The results indicated that Facterra successfully engaged players and conveyed its intended message, further showing that games indeed can be used to raise awareness. However, feedback also highlighted areas for improvement, particularly in terms of polish, depth, and clarity of educational elements
Natural Language Processing and Large Language Models for Automation of Compliance Tracing
No full textCompliance is a costly and time-consuming task that most, if not all, firms must perform. As such, automating parts of the compliance process could be highly valuable. This thesis aims to investigate challenges faced by European software-intensive firms in their compliance processes, identify automation opportunities, and develop a Natural Language Processing- and Large Language Model-based software artifact to automate compliance tracing between company guidelines and normative requirements. The thesis followed the Design Science Research approach, and as such, the research was conducted in close collaboration with industry practitioners. The challenges and automation opportunities were identified together with seven interviewees from four different companies, and the final software artifact, dubbed TraceAlign, was developed and evaluated in focus groups with a total of twelve unique participants from two companies. The identified challenges ranged from organizational- and management-related to specifics inherent to the specifications of normative requirements. Automation opportunities related mainly to the management of requirements, company guidelines, and compliance evidence, of which this thesis focuses specifically on the task of compliance tracing of company guidelines to normative requirements. The final software artifact, TraceAlign, was considered to be time- and cost-saving by the focus group participants, but could perhaps be made more accurate. We conclude that there are many challenges with compliance that could potentially be automated using Natural Language Processing and Large Language Models