This project is centered around digital twins, and the application of formal methods for predictive maintenance and anomaly detection for mobile concrete batching plants (MCBP).

Sensors will be placed within the MCBP to collect real-time data, such as operational and performance metrics for monitoring. This would enable us to make data-driven decisions and state predictions, thus bridging the gap between the physical and digital systems.

Anomaly detection is driven by data, analytics, and machine learning. These algorithms continuously monitor the data for deviations from defined patterns. This is a proactive approach that helps minimize downtime and optimizes the maintenance and operational efficiency of the MCBP.

In summary, the project will use technology to transform a manual industry to become digitalized, thus enhancing efficiency and reducing costs for both the production and delivery of the concrete and the MCBP itself.

Modern robotic systems must adapt to drastic and unprecedented changes in their environment at runtime, i.e., they must be self-adaptive.

The RoboSAPIENS project, which my PhD is a part of, explores creating an architecture for trustworthy self-adaptation. This entails that the system successfully adapts to new environments whilst maintaining safe behavior, which is not a trivial task.

On one hand, the exact behavior of a self-adaptive system is difficult to determine, as it changes depending on the environment it is operating within. On the other hand, robotics are typically verified as being safe through certifications based on their exact behavior. Thus, we need to find a way to ensure that the robot still lives up to its safety requirements despite changing its behavior to accommodate changes to the environment.

My research will be aimed at contributing towards this objective through three main tasks:

1. Domain Analysis and Ontology Development: Collaborate closely with industrial partners to create an ontology capturing the commonalities and differences in verification requirements across domains.
2. Language Definition and IDE Development: Define domain-specific language(s) to describe safety procedures within adaptation loops.
3. Formally Verified Trustworthiness Checkers: Generate formally verified trustworthiness checkers and accompanying documentation. 

My research work focuses on contributing to the Tactile Internet. Tactile Internet is the foundation of the use cases such as Teleoperation, and Immersive Virtual Reality on the table. The key to the Tactile Internet is its ability to transmit information quickly, ensuring that users interacting with devices on the network feel like it's happening in real-time.

Different strategies are proposed to provide the users with an excellent experience in interacting with endpoints. One of which is called “distributed intelligence”. To minimize the latency, local edge devices are organized to work together to offload the computation remote cloud. In that way, the workload on the cloud is reduced thus improving overall performance.

Through this localized intelligent processing, the goal is to make user interactions smoother and more natural.

Extensive screening and long-term home monitoring programs have proven effective in reducing the impact of diseases. The introduction of screening pregnant women and initiating medical treatment for high-risk women holds the potential to mitigate the effects of pregnancy-related hypertensive disorders. Annually, these hypertensive disorders contribute to half a million maternal deaths worldwide. However, the expenses associated with implementing such screening are currently hindering the adoption in hospitals, with staff resources emerging as a critical factor.  

This project aims to enhance healthcare quality, efficiency, and cost-effectiveness in hospital and home settings by implementing sensor-based self-measuring stations, monitoring support tools, and personalized medicine. The ultimate goal is to establish automated screening, digital coaching, and decision support tools to assist in screening and initiating personalized medicine to mitigate the impact of pregnancy-related hypertensive disorders.

Automatic sleep scoring make use of algorithms inferring sleep states from sleep recordings. In recent years, these algorithms have shown great performance, even reaching human performance, by employing various deep learning approaches. Most of these models are trained on recordings of healthy patients or a few specific patient groups, and they can not be expected to generalize to other patient groups as sleep differs with both age and health. In addition, sleep EEG recordings are expensive to produce and label for training.

In this project, we will seek to develop and test “transfer learning” or “semi-supervised learning” approaches to this problem, adapting high performing models trained on large data sets to perform well on smaller data sets recorded for different patient groups or with a different device.

In particular, the developed methods will be used not only for regular, clinical sleep recordings, but also for recordings made using the ear-EEG device developed at the Center for Ear EEG.

This project is carried out with close collaboration with the Danish companies Vestas, Velux, Millpart, Hydrospecma, and Landia, and the Ringkøbing-Skjern municipality.

Each company focuses on different products and industries. The goal of this project intends to contribute to digital technologies of the involved companies for improving product design and production using virtualization and simulation methods, considering strategies to enhance optimization and efficiency of the processes.

This will be achieved through the combination and integration of already existing tools for digital twin representation and software for process and product management. This PhD project is associated with the Digital Transformation Lab (DTL) in Ringkøbing-Skjern Municipality targeting a digital transformation of the 5 companies in the area. The DTL will be used for the experimental parts of this PhD project. The DTL will be key in ensuring the best exchange of research-based knowledge from the university with practice-based knowledge at local companies.

Auditory Attention Decoding (AAD) is envisioned to become an important technology for the next generations of hearing aids. With an AAD component, a hearing device will be able to detect which sound sources the user is attending, and thereby the audio signal processing in the hearing device can be adapted accordingly.

In this project, we propose a new approach in which the AAD is based on cognitive processing of the audio event stimuli. This is expected to better separate attended and non-attended sound sources. The approach try to extract event related potentials and analyze using state-of-the-art deep learning model. We will validate the algorithm on Ear-EEG signal which is minimalistic and will allow the  technology to be implemented in future hearing aid devices.

The ReMaRo ETN project aims to develop Artificial Intelligence methods for  submarine robotics with quantified reliability, correctness specifications, models, tests, and analysis & verification methods.

Within this project, I am involved in the development of deep learning algorithms for vision-based navigation for underwater safety critical applications.  One of the underwater contexts contemplated within the projects is pipeline inspection. That is, I will develop visual (camera-based) localization algorithms for pipeline inspection, supported by other sensors such as sonars. Moreover, I will work on the reliability assessment of these methods with the help of other PhD students from the ReMaRo consortium to identify whether if the localization measurements are reliable or not depending on the quality of the measurements taken.