Chalmers Open Digital Repository
Not a member yet
26247 research outputs found
Sort by
Designing a Proactive Security Solution for Kubernetes Clusters
No full textKubernetes cloud environments often contain security vulnerabilities exploitable during runtime, but undetected by the scanning tools widely used during build or deployment phases, due to the distributed and dynamic nature of containerized
applications. Traditional reactive security approaches for detecting runtime exploits usually depend on manual intervention, causing delayed responses and unnecessary service disruptions because of false positives.
This thesis develops and evaluates a proactive security solution for Kubernetes clusters, with particular focus on 5G core network environments. The proposed solution integrates multi-step attack detection with automated mitigation strategies
to improve detection accuracy and response time. Using a labeled dataset of Falco alerts generated within a 5G core network by the ACE-WARP project, five statistical and machine learning models are developed and evaluated: Markov Chain, Hidden Markov Model, Long Short-Term Memory (LSTM), Convolutional Neural Network (CNN), and Bayesian Network. With each model, the likelihood of an attack sequence belonging to the normal or attack distributions is used for the binary classification task of identifying sequences containing attacks. All models are evaluated with both balanced (1:1) and imbalanced (1:50) class ratios. The LSTM and CNN models achieve the highest detection rate of 84.3 percent and the lowest false alarm rate of 0.02 percent under imbalanced settings. They maintain strong precision (average precision of 99.2 percent) and area under the curve (AUC-ROC of 96.2 percent), outperforming the baseline Markov Chain, other
models, and prior work. Beyond identifying the best-performing models, this thesis also proposes mitigation techniques such as pod deletion for immediate isolation of threats and sandboxing for forensic analysis. These components form a complete security solution designed as a containerized microservice using Docker and exposed through a RESTful API, that can be applied to large Kubernetes clusters. Therefore, this design supports scalability, modularity, and compatibility requirements with industrial 5G core network deployments
Catalytic Activity of Amyloids of α-Synuclein Variants with Pathological Mutations
No full textParkinson’s disease is a progressive neurogenerative disorder which is characterized by the presence of Lewy bodies, which consist of insoluble α-Synuclein amyloid aggregates. While amyloids have been considered to be inert end-products, recent studies show that some amyloids exhibit catalytic properties. This master’s thesis focuses on the catalytic activity of α-Synuclein, including the wild-type and three pathological mutations: E46K, H50Q and A53T. In this thesis, the monomers of each variant were over-expressed in Escherichia coli, purified, aggregated as well as quality controlled through circular dichroism, Thioflavin T fluorescence and atomic force microscopy. The catalytic activity of these four amyloid fibril variants were studied in relation of their control group consisting either monomers or buffer background. Substrate paranitrophenyl acetate was used to study esterase activity, while paranitrophenyl phosphate and adenosine 5’-triphosphate were used for dephosphorylation assays. Results show that all variants exhibit strong esterase activity, while phosphatase activity varies between the variants and individual experiments, suggesting sensitivity on how amyloid fibrils are prepared. The dephosphorylation results are discussed in terms of how fibril polymorphism may influences the catalytic activity. While esterase activity was high, the differences between the variants were not statistically significant enough to draw firm conclusions
Investigating Project and Asset Management Integration Ensuring Reliable Handovers in Production Projects
No full textEfficient management of resources, ensuring reliable operations, has become increasingly important for Swedish Match. A critical aspect of achieving this is the transition from project
development to implementation. Challenges often arise during the handover due to unclear processes, missing information, and inadequate planning, leading to inefficiencies.
This thesis analyses project structures and responsibilities. Identifying problem areas in the different project phases, the thesis aims to find ways to ensure operational reliability and
effective data management. The purpose is to find critical areas in the project process that hinder production and operation & maintenance from managing the production equipment over its lifecycle.
The thesis addresses key challenges in the handover of production projects concerning project management and asset management integration and investigates solutions that can be implemented.
The methodology includes a literature study, a case study, and a process study. The case study involves analyses of documented processes and structures, alongside interviews. The process study supports the findings
by following an ongoing project and gathering information through documentation analysis and observations. The analyses identified several areas of improvement and challenges concerning project handover.
These include the necessity of clear roles and responsibilities. Setting higher standards by the steering committee, overseeing prioritization, and continuously reviewing ongoing and previous projects. Involvement of the receiving organizations in project orders was identified as important, together with detailed handover requirements to reach satisfaction from both parties. Establishing a standard for project documentation was also identified as necessary, alongside ensuring effective and reliable communication channels
Road Friction Aware Adaptive Cruise Control using Robust Nonlinear Model Predictive Control with Uncertainty Quantification
No full textRoad friction is a fundamental factor affecting vehicle safety, especially under adverse
weather conditions such as rain, snow or ice. Reduced road grip increases
stopping distances and the likelihood of loss of control, contributing to a significant
number of traffic accidents each year. While modern Advanced Driver Assistance
Systems (ADAS) are designed to enhance driving safety, many systems still lack the
ability to adapt their behaviour in real time to varying road friction levels, limiting
their effectiveness in low-friction or rapidly changing weather conditions. This
thesis addresses this limitation by developing a friction-aware, curvature-adaptive
Adaptive Cruise Control (ACC) framework based on a Robust Nonlinear Model
Predictive Control (NMPC) formulation.
The controller anticipates road curvature and spatially-varying friction by incorporating
predicted profiles into the optimization problem. To account for uncertainty,
friction is represented through position-dependent bounds in the MPC, while
stochastic realizations of friction are generated via Beta sampling and the lead vehicle’s
acceleration varies randomly within feasible physical limits. These variations
are reflected in dynamic safety constraints that adapt to evolving conditions ahead
of the ego vehicle.
Soft constraint relaxation is introduced to maintain feasibility under conflicting demands
such as sudden friction drops or aggressive lead vehicle deceleration. Despite
the nonlinear nature of the problem and the presence of uncertainty, the controller
operates using an efficient CasADi-based implementation in MATLAB.
Simulation results demonstrate that the proposed framework achieves robust, adaptive
cruise control under friction and road curvature changes with stochastic disturbances.
The integration of environmental awareness into predictive control enables
safer, more responsive vehicle behaviour without sacrificing passenger comfort, and
highlights the feasibility of embedding the friction information into future ADAS
systems
Phase Control Implementation in an UWB Hyperthermia Prototype System
No full textThis degree project investigates the implementation of phase regulation in a prototype
ultra-wideband (UWB) hyperthermia system developed at Chalmers. A
proportional control algorithm was integrated using an AD8302 gain and phase
detector, with calibration performed to relate voltage outputs to known phase differences.
The system was tested in an emulated feedback loop using MATLAB and
Arduino to control digital phase shifters. Testing at 650 MHz demonstrated optimal
performance, while results at 300 MHz and 900 MHz revealed incomplete or irregular
phase shifts. Although verification relied on visual inspection due to limited
measurement equipment, the regulator successfully adjusted signal phases towards
defined targets. Further improvements are required to achieve higher precision and
system reliability
Algorithmic Optimization of Warehouse Box Selection
No full textThe efficiency of e-commerce logistics chains is heavily influenced by bin selection
during the packing process, which can be formulated as a variant of the NP-hard
three-dimensional bin packing problem: selecting the smallest possible combination
of heterogeneous bins that can fit a given order. This thesis presents the implementation
and evaluation of several solving techniques for this problem, with a focus
on achieving runtimes suitable for real-world applications. Following a survey of
what solutions currently exist in the literature, the four commonly found methods
are heuristics, Genetic Algorithms (GAs), Deep Reinforcement Learning (DRL),
and Mixed-Integer Linear Programming (MILP) solvers. These are all evaluated
throughout this thesis, as well as the CP-SAT solver, which is capable of handling
MILP problem formulations. Contrary to common findings within the literature,
which consider GAs and DRL methods as state-of-the-art, this thesis shows that
when replacing the traditional MILP solvers with CP-SAT, this becomes a viable
alternative that can provide increased packing efficiency in a feasible amount of time.
Heuristic solutions can also be a suitable option in certain scenarios, for instance,
when minimizing computational load is highly prioritized.
This thesis concludes that it is possible to generate solutions to this bin packing
problem using CP-SAT within feasible runtimes. While human workers may outperform
the solver in some cases, this is primarily due to limitations within the
current model. The solver is ready to be implemented if a warehouse meets a few
key conditions: items should be well approximated by rectangular prisms, and the
input data must be accurate and reliable. Given the NP-hard nature of the problem,
performance is better when the number of items remains within normal operational
bounds. Additionally, the availability of appropriately sized packing cartons is important.
Without this, the benefits of the bin selection automation are significantly
reduced
Multikriterieanalys för olika blågröna lösningar inom kvartersmark - Med Göingegården i Varberg som fallstudieområde
No full textMed en ökad risk för mer frekvent förekommande extremväder till följd av
klimatförändringar krävs åtgärder för att anpassa samhället till de förändringar som
sker. Blågröna lösningar är exempel på åtgärder som förebygger översvämningar och
samtidigt bidrar till ökad biologisk mångfald. Lagstiftning och direktiv gällande
hållbarhet är högst relevanta idag där allt från FN:s utvecklingsprogram (UNPD),
Europaparlamentet och svensk lagstiftning innehåller information om hållbar
dagvattenhantering. Detta är även aktuellt för Varbergs kommun där bland annat ett
nytt bostadsområde, Göingegården, planeras med flera ingående blågröna lösningar.
Varbergs kommun understryker i sin översiktsplan vikten av att anpassa kommunen för
framtida konsekvenser av klimatförändringar.
Syftet med detta arbete var att fastställa vilken blågrön lösning som är bäst lämpad inom
kvartersmark, med avseende på biologisk mångfald och dagvattenhantering. Där
Göingegården i Varberg tillämpas som fallstudieområde. Den metod som tillämpas för
att utvärdera blågröna lösningar är en så kallad multikriterieanalys (MKA), där olika
lösningar jämförs utifrån samma kriterier. Utöver MKA genomfördes även en
inledande litteraturstudie för att samla referenser med relevant information. För att
avgränsa arbetet utvärderades studien tre valda dagvattenlösningar: gröna tak,
regnvattentunnor och regnbäddar.
Resultatet av studien visade att regnbäddar är bäst lämpade för både biologisk mångfald
och dagvattenhantering, medan regnvattentunnor tilldelas näst bäst resultat och lägst
resultat får gröna tak. Ett alternativ till arbetet är att använda en annan MKA-modell
som ger andra förutsättningar, vilket kan skapa en bredare grund för studien och leda
till ett mer tillförlitligt resultat. En möjlig vidareutveckling av arbetet är att använda ett
modelleringsverktyg för att tillämpa de blågröna lösningarna i praktiken
Fan placement on a Formula Student vehicle: A study to evaluate placement of fans for improved aerodynamic performance
No full textFormula Student 2024 saw an increase in teams that used fans for aerodynamic gain instead
of just for cooling purposes. This gave a huge performance to the car in terms
of downforce and therefore better grip which allows for faster cornering. The sudden
performance gain resulted in a rule change for the 2025 Formula Student season where a
maximum combined total rated power of any active devices designed to move air is 500
W. This restriction made it more difficult to find aerodynamic advantages from the fans
and their positioning became more vital.
The objective for this project is to investigate how a fan system on Chalmers Formula Student’s
(CFS) current car can be implemented to enhance the aerodynamic performance.
Concepts were generated by dividing the aerodynamic components into four regions, front
wing, side wing, rear wing and diffuser. Each component could then be investigated by
looking at data from past simulations, but also by taking inspiration from other teams.
Having formulated several concepts for each geometrically defined region, they were realized
using computer aided design (CAD). The results were then evaluated using computational
fluid dynamics (CFD) and numerical values and field plots of several properties
were obtained. The result showed that significant downforce can still be found in spite of
the new regulation change. The regions that performed best were side wing and diffuser.
While the concepts in these regions show potential, future iterative work is still needed
to perfect their functionality, depending on the direction of development for the CFS
aerodynamic package
Hydrogeologisk studie av multifunktionella blågröna tak
No full textI takt med en ökad urbanisering och klimatförändringar blir behovet av hållbara
lösningar för dagvattenhantering alltmer angeläget. Detta arbete utreder potentialen
hos blågröna tak som en innovativ metod för att hantera regnvatten, minska
översvämningsrisk och bidra till en mer resilient stadsmiljö. Studien fokuserar på en
blågrön taklösning i Göteborg som en fallstudie, med syfte att analysera hur dessa
system kan bidra till lokal dagvattenhantering, energieffektivitet och ekologiska
värden.
Arbetet kombinerar en litteraturstudie med praktiska mätningar av nederbörd,
temperatur och vattennivåer under en period på våren. Resultaten visar att det
blågröna taket kan magasinera och fördröja regnvatten effektivt, vilket avlastar
stadens avloppssystem vid nederbörd. Dessutom påvisades en viss isolerande effekt
i takkonstruktionen, samt potential för ökad biologisk mångfald.
Trots begränsningar i form av låg nederbörd under mätperioden visar studien att
blågröna tak har stor potential som del av framtidens hållbara stadsplanering. För att
fullt ut utvärdera systemets kapacitet rekommenderas vidare studier över längre
tidsperioder och under varierande väderförhållanden. Slutsatserna i denna rapport
kan fungera som ett underlag för kommuner, fastighetsägare och andra aktörer som
vill implementera mer hållbara lösningar i urbana miljöer