Dennis Joosens

• 21 september 2026
• 16.00 uur 
• ​UCSIA-huis, Manresazaal
• Promotoren: prof. dr. Maarten Weyn & prof. dr. Rafael Berkvens 
• Faculteit Toegepaste Ingenieurswetenschappen

Abstract

Wireless communication technologies such as Wi Fi, Bluetooth, 4G, and 5G have become fundamental to modern society, enabling pervasive connectivity through radio-frequency (RF) signals. Beyond communication, these signals can also be exploited for environmental sensing. This dissertation investigates whether crowd presence and density can be estimated non-intrusively by analysing the aggregate energy of ambient RF signals, without decoding transmitted data or collecting privacy-sensitive information. A measurement framework is developed to acquire and analyse RF signals in both the time and frequency domains, with emphasis on computationally efficient signal-processing techniques and waveform characterization. Experimental campaigns conducted in diverse environments demonstrate that spectral activity metrics correlate with occupancy levels and can distinguish densely populated settings from sparse ones. However, the results show that spectrum measurements alone do not uniquely determine crowd size, as they are also influenced by device usage patterns, application behaviour, transmission activity, and protocol dynamics. Consequently, ambient spectrum analysis is best viewed as a coarse-grained, protocol-agnostic sensing modality that complements emerging integrated communication and sensing systems, while offering promising opportunities for future research in intelligent spectrum monitoring and occupancy estimation.

Wouter Jansen

• 10 september 2026
• 17.00 uur 
• ​Campus Middelheim, lokaal m.A.143
• Promotor: prof. dr. Jan Steckel 
• Faculteit Toegepaste Ingenieurswetenschappen

Abstract

Modern autonomous robots rely heavily on optical sensors. In harsh industrial environments filled with dust, fog, or smoke, these sensors quickly fail, making in-air ultrasonic acoustics a highly robust alternative. While such 3D in-air sonar sensors already excel at providing geometric and spatial awareness, they lack contextual and cognitive understanding. We know this rich semantic information is present in broadband acoustic signals because we clearly observe it in the complex echolocation behavior of bats.
This thesis presents a framework to get closer to a true semantic understanding for robotic acoustic perception. First, a high-fidelity simulation environment was developed to test robotic perception across different sensor modalities. Next, we engineered a ruggedized 3D sonar platform capable of operating in real-time under severe industrial conditions. Finally, this work explores several advanced applications to push the boundaries of acoustic autonomy. This includes fusing sonar with motion sensors to stabilize perception on rough terrain, implementing multi-sensor acoustic navigation strategies, and applying deep learning to raw acoustic echoes to reliably classify objects with unique acoustic signatures. Ultimately, this work addresses several challenges and provides a practical pathway to more robust autonomous robots to navigate and interpret the most challenging environments.

Anthony Schenck

• 7 september 2026
• 16.00 uur 
• ​Campus Middelheim, lokaal m.A.143
• Promotoren: prof. dr. Walter Daems & prof. dr. Jan Steckel 
• Faculteit Toegepaste Ingenieurswetenschappen

Abstract

Automating the detection process of compressed air leaks in industrial environments can have a significant impact on wasted energy. The methods currently employed to detect these leaks are both costly and time consuming, while their costs are hidden in the total cost of energy.

With AirleakSLAM we attempt to automate the detection and locailization process, providing quick and accurate results with (almost) no manual labor required, even in the largest industrial environments.

Seyed Ali Golmohammadi Tavalaei

• 3 september 2026
• 15.30 uur 
• ​Campus Middelheim, lokaal m.G.010
• Promotor: prof. dr. David Hernando & dr. Navid Hasheminejad
• Faculteit Toegepaste Ingenieurswetenschappen

Abstract

Modern road infrastructures require continuous monitoring to ensure their safety, durability, and efficient maintenance. Traditional inspection methods are often costly, disruptive, and provide only limited information about pavement condition. Embedded fibre optic sensors offer a promising alternative by enabling continuous, real-time monitoring of structural behaviour. However, the large volume of data generated makes automated analysis essential. This dissertation presents an intelligent structural health monitoring framework that combines embedded Fibre Bragg Grating (FBG) sensor networks with artificial intelligence to assess pavement condition automatically. The proposed approach integrates sensing, data processing, and machine learning to transform raw sensor measurements into meaningful indicators of structural health. The framework employs unsupervised learning techniques for anomaly detection and deep learning models to identify long-term structural trends and generate compact health indicators. Signal processing and neural network models are used to reduce data volume while preserving information related to traffic loading, environmental effects, and pavement response. The proposed methods are validated using laboratory experiments, field measurements from instrumented road sections, and simulated datasets. The results demonstrate that continuous sensing data can be efficiently processed into reliable and interpretable indicators for long-term pavement monitoring. Overall, this research contributes to the development of intelligent, data-driven road infrastructure by providing an automated and scalable framework for proactive pavement monitoring, supporting more effective maintenance planning, reduced lifecycle costs, and improved infrastructure resilience.

Milan Cornelis

• 29 juni 2026
• 17.00 uur 
• ​Campus Middelheim, lokaal m.A.143
• Promotor: prof. dr. Paul De Meulenaere 
• Faculteit Toegepaste Ingenieurswetenschappen

Abstract

Modern technological products such as robotic vacuum cleaners, self-driving cars, or surgical robots are designed by engineering teams from different disciplines: mechanics, electronics, software, and so on. Systems like these, where all those disciplines come together, are called mechatronic systems.
The problem is that these components strongly influence one another. If a team decides to replace a motor with a more powerful one, this has consequences for the battery, the gears, and the control system. If the other teams are unaware of this, the final product may fall short or even fail entirely. Such mistakes only become visible late in the development process, resulting in longer development times and higher costs.
To speed up the design process, engineering teams often work independently and in parallel. However, this makes it even harder to maintain an overview of the system as a whole. Existing tools and methods do offer partial solutions, but no complete answer to this problem.
This thesis therefore proposes a knowledge model that explicitly tracks how design decisions made by one team affect the other teams. The goal is for everyone involved to know at any point what they can expect from one another, so that the design remains consistent and collaboration runs more smoothly.

Barbara Bohlen

• 18 juni 2026
• 17.00 uur 
• ​Campus Drie Eiken, lokaal d.O.06
• Promotor: prof. dr. Tom Breugelmans 
• Faculteit Toegepaste Ingenieurswetenschappen

Abstract

To combat the global climate crisis, we must transition towards a circular carbon economy. This research addresses a major bottleneck in current capture technologies: the energy penalty required to release CO₂ from liquid capture solutions, which typically demands massive amounts of heat. By using electrochemistry, this thesis demonstrates how electricity and specialized metal catalysts can bypass this heat-intensive step to either transform captured CO₂ directly into valuable industrial chemicals, such as formate and acetate, or efficiently release it for further use. This work provides a strategic framework for the integration of carbon capture and utilization. Ultimately, these findings move us closer to a future where CO₂ is no longer just a waste product to be stored, but a sustainable raw material for the fuels and chemicals of the future.

Dorothee Goettert

• 10 juni 2026
• 15.00 uur 
• ​Stadscampus, Klooster van de Grauwzusters, Promotiezaal
• Promotor: prof. dr. Jan Dries 
• Faculteit Toegepaste Ingenieurswetenschappen

Abstract

Every time a potato is processed into fries or chips, large volumes of water are used, and this potato processing waster needs to be treated before it can be safely discharged into the environment. A common treatment process in the potato processing industry is the use of microbiology for Nitrogen and Carbon removal, but the addition of chemicals for Phosphorus removal. These metal salts bind Phosphorus and pull it out of the water into the sludge, which is then regularly wasted, thus removing Phosphorus out of the system. It works, but it comes with a cost: ongoing chemical purchases, cost of sludge disposal, and phosphorus so tightly bound it can become difficult to recover and reuse.
Biology offers an interesting alternative through specialized bacteria known as "polyphosphate accumulating organisms" or PAOs. Instead of using chemicals to trap phosphorus in a permanent bond, these microbes, under the right environmental conditions, absorb and store large amounts of phosphorus within their own cells as a natural energy reserve. Because the microbes hold the phosphorus in a concentrated, organic form rather than locking it away with metal salts, it’s possible to recover these from the wasted sludge.
As Phosphorus is a finite and essential resource for global food security, the central question driving this research was whether we could do away with that chemical dependency entirely, and replace it with microbiology. Could we harness that capacity in an industrial setting, treating the complex carbon and variable wastewater that a potato processing plant produces day in, day out? This dissertation follows these questions from the laboratory bench all the way to a working industrial installation. It investigates which microbial communities are needed, how to coax them into existence from ordinary activated sludge, what operational conditions allow them to thrive, and what happens when the carefully controlled world of the lab meets the reality of full-scale industrial operation.

Zhaoxing Wang

• 9 juni 2026
• 16.00 uur 
• ​Campus Middelheim, lokaal m.A.143
• Promotors: prof. dr. David Hernando & prof. dr. Zhi Cao 
• Faculteit Toegepaste Ingenieurswetenschappen

Abstract

Roads are everywhere around us, quietly supporting our daily lives by connecting people and transporting goods. Yet behind this convenience lies a significant environmental cost: building and maintaining roads requires vast amounts of raw materials and contributes to climate change. In many ways, our road networks are not just infrastructure, but also large reservoirs of resources waiting to be reused.

This PhD research explores how roads can be made more sustainable by rethinking how these materials are used. It focuses on recycling existing road materials and introducing plant-based alternatives, turning what is often considered waste into valuable resources. However, making this transition is not straightforward. It requires understanding where materials are located, how reliable environmental estimates are, how impacts change over time, and how decisions affect entire road networks.

By combining data, models, and real-world conditions, this research demonstrates that smarter use of recycled materials can significantly reduce both emissions and costs. Ultimately, it offers new approaches to designing and managing roads that are not only functional, but also more sustainable for the future.

Sven Arnouts

• 18 mei 2026
• 17.00 uur 
• ​Campus Drie Eiken, lokaal d.O.5
• Promotors: prof. dr. Tom Breugelmans, prof. dr. Sara Bals & dr. Thomas Altantzis 
• Faculteit Toegepaste Ingenieurswetenschappen

Abstract

This work applies a case-study driven approach to unravel how the structure of metal-based catalysts affects their performance in the electrochemical reduction of CO₂. Through a multimodal approach, it was examined how these catalysts change during operation and how these transformations influence efficiency and stability. Gaining insight into this relationship helps guide the design of improved catalysts, bringing electrochemical CO₂ conversion closer to sustainable industrial application.

Alana Rossen

• 6 mei 2026
• 17.00 uur 
• ​Campus Drie Eiken, lokaal d.Q.002
• Promotor: prof. dr. Tom Breugelmans 
• Faculteit Toegepaste Ingenieurswetenschappen

Abstract

The rapid rise in atmospheric carbon dioxide CO₂ concentrations necessitated by anthropogenic activity requires the development of robust carbon capture and utilization (CCU) technologies. Among these, electrochemical CO₂ reduction (CO₂RR) offers a promising pathway to synthesize value-added chemicals and fuels using renewable energy. While laboratory research typically occurs at ambient conditions, industrial-scale electrolyzers will inevitably operate at elevated temperatures due to internal ohmic heating. This dissertation investigates the influence of thermal intensification on the performance, stability, and transport phenomena of electrochemical CO₂ and carbon monoxide (CO) reduction.
One major challenge at 85°C is the morphological degradation of catalysts. To address this, a bismuth-based catalyst integrated with a mesoporous carbon shell was developed. This architecture effectively suppressed bismuth sintering and surface degradation, maintaining high selectivity toward formate for twenty-four hours of continuous operation. Beyond catalyst stability, elevated temperatures exacerbate gas diffusion electrode (GDE) flooding, which disrupts the triple-phase boundary necessary for efficient gas conversion. By implementing precise differential pressure control between the gas and liquid phases, the reaction interface was stabilized, increasing the Faradaic efficiency for formate from 40% to 65% under thermally intensified conditions.
The research further transitioned to electrochemical CO reduction within a zero-gap electrolyzer configuration. This architecture circumvents the carbonate salt precipitation issues inherent to CO₂ feeds, allowing for a clearer decoupling of thermal effects from bulk electrolyte chemistry. This study revealed that while temperature has a secondary effect on the intrinsic product distribution on copper catalysts, it drastically reduces the crossover of liquid products through the ion-exchange membrane. This reduction in crossover is a critical finding for industrial process design, as it simplifies downstream separation and enhances product recovery.
Ultimately, this work demonstrates that elevated temperature operation fundamentally alters the electrochemical environment, influencing parameters ranging from CO₂ solubility and local pH to membrane permeability and catalyst longevity. By exploring the interplay between catalyst design, mass transport, and reactor architecture, this dissertation provides a comprehensive framework for the design of next-generation electrolyzers. These insights are essential for the deployment of thermally intensified systems capable of operating at the scales required for global carbon mitigation.

Stuti Pathak

• 1 april 2026
• 16.00 uur 
• ​Campus Middelheim, lokaal m.A.143
• Promotor: prof. dr. Rudi Penne & prof. dr. Gunther Steenackers 
• Faculteit Toegepaste Ingenieurswetenschappen

Abstract

In this defence, we explore how non-Euclidean machine learning can transform raw point clouds into clean, reliable, and high-quality representations for a wide range of engineering domains. In particular, we present novel methods to efficiently simplify large datasets, recover missing information, and reconstruct smooth surfaces from noisy inputs. Together, these contributions bring 3D vision closer to robust and real-time applications in everyday life.

Marthe Nees

• 4 maart 2026
• 15.30 uur 
• ​Stadscampus, Klooster van de Grauwzusters
• Promotor: prof. dr. Christophe Vande Velde & prof. dr. Pieter Billen 
• Faculteit Toegepaste Ingenieurswetenschappen

Abstract

This thesis investigates new ways to chemically recycle polyurethane (PU), with special attention to the isocyanate-based part of the material, which is often overlooked. While most recycling research focuses on recovering polyols, this work shows that the isocyanate fraction can also be reused effectively in a circular production process. An improved alcoholysis method was developed to break down PU into carbamates while preventing the formation of unwanted amines. These carbamates were then converted back into valuable isocyanates through thermolysis. The study also explored process optimization and an alternative application as glue for the mechanical recycling of PU. Overall, the work demonstrates a practical and flexible pathway toward more sustainable and circular PU recycling.

Michiel De Rop

• 5 maart 2026
• 17.00 uur 
• ​Stadscampus, lokaal s.C.002
• Promotor: prof. dr. Jonas Hereijgers 
• Faculteit Toegepaste Ingenieurswetenschappen

Abstract

This thesis explores the use of structured 3D electrodes in electrochemical flow reactors to improve performance while keeping energy losses low. In order to quantify this equilibrium, a new evaluation metric, the Hydrodynamic Electrode Performance Factor (HEPF), was introduced. This was needed because conventional electrode designs often focus on maximizing surface area, while neglecting pressure losses that limit overall reactor efficiency and increase operating costs. Through experiments and simulations on pillar-array and 3D-printed TPMS electrodes, the study demonstrates the validity of the newly introduced metric and shows that an engineered electrode design can outperform the surface area alone.

Cristina Moyaert

• 12 februari 2026
• 16.00 uur 
• ​Stadscampus, Hof van Liere, F. De Tassiszaal
• Promotoren: prof. dr. Pieter Billen & prof. dr. Philippe Nimmegeers 
  
• Faculteit Toegepaste Ingenieurswetenschappen

Abstract

This thesis introduces new metrics to evaluate a pressing issue in the transition to plastics circularity; their increasing complexity. At its core, circularity largely deals with managing this complexity, which necessitates its quantification. Using statistical entropy across different levels, i.e. molecular, product and geospatial levels, can support product design and waste management decision-making.

Rens Baeyens

• 6 februari 2026
• 16.30 uur 
• ​Campus Middelheim, lokaal m.A.143
• Promotoren: prof. dr. Walter Daems & prof. dr. Jan Steckel   
• Faculteit Toegepaste Ingenieurswetenschappen

Abstract

Embedded systems play a key role in collecting data from sensors in modern applications. However, existing data-acquisition solutions are often tailored to specific use cases, making them difficult to reuse or adapt.
This dissertation presents a flexible embedded data-acquisition framework that can be applied across different applications. The framework combines modular system design with efficient data compression and accurate time synchronization, enabling reliable and efficient handling of sensor data on resource-constrained devices.
The proposed approach is validated through practical implementations on embedded platforms. The results demonstrate that flexible, scalable, and application-independent data-acquisition systems are achievable, supporting both industrial and research use cases.

Boris Gilis

• 2 februari 2026
• 16.00 uur 
• ​Campus Drie Eiken, lokaal d.Q.002
• Promotor: prof. dr. Iris Cornet   
• Faculteit Toegepaste Ingenieurswetenschappen

Abstract

What if we could use yesterday’s cooking oil to replace fossil oil in the production of valuable chemicals? Despite growing sustainability efforts, the chemical industry still relies heavily on fossil resources and energy-intensive production methods. Meanwhile, large amounts of cooking oil and fat wastes are used as a low-value energy source, even though they contain valuable building blocks for chemical production.
This PhD research investigated whether waste oils and fats could be converted into long-chain dicarboxylic acids. These acids are versatile components used in materials such as plastics and coatings. Producing these molecules efficiently with conventional chemical processes is challenging. Instead, this research explored a more sustainable approach by using yeasts as tiny biochemical production factories.
By developing an optimised and controlled fermentation process, these yeasts converted the waste oils and fats into high amounts of long-chain dicarboxylic acids. This work shows that even complex waste streams can be transformed into valuable products, reducing waste and supporting the transition of the chemical industry towards renewable and circular raw materials.