Modern factories showlittle resemblance to the assembly lines froma century ago. Nowadays, a single human might be responsible for setting up, operating, and maintaining a complex chain of machines. This shift has made the work of factory operators more cognitively demanding, requiring constant monitoring and problem-solving. Key to tackling these challenges is the effective collaboration of human operators in exchanging ideas and knowledge, from high-level problem-solving strategies to solutions for emerging issues. Yet, the heightened productivity requirements and small shifts mean fewer opportunities to share this knowledge face-to-face and reporting practices are deprioritized. Thus, the manufacturing industry faces a knowledge management crisis.
This dissertation investigates the integration of conversational AI assistants into manufacturing settings to facilitate knowledge sharing among factory operators. Capitalizing on recent advancements in Natural Language Processing (NLP), particularly in Large Language Models (LLMs), this research investigates the designing and evaluation of conversational AI tools that efficiently capture and share human knowledge on the factory floor while addressing operator needs and concerns. The introduction of conversational AI assistants for knowledge sharing—referred to as cognitive assistants (CA) in this work—in factory environments promises significant benefits but comes with numerous challenges....@en

In the rapidly evolving healthcare landscape, integrating Artificial Intelligence (AI), particularly Large Language Models (LLMs), presents significant opportunities and complex challenges. This study examines the efficacy and implications of LLMs within healthcare systems, with a focus on enhancing user-centric evaluation methods for AI-generated healthcare advice. The research centres on three critical areas: Identifying the limitations of current evaluation methods, addressing the challenges related to the accuracy and reliability of LLM-generated advice, and proposing improvements to evaluation frameworks to enhance the practical application of these models in healthcare settings.

Our study utilizes a chatbot prototype trained specifically on healthcare datasets relevant to the Dutch context, exploring its application in real-world scenarios to validate and refine evaluation metrics. By involving healthcare professionals in interactions with the chatbot, we aim to ground our findings in practical, user-based experiences. The engagement with the prototype helps uncover vital insights into the AI’s performance, emphasizing the necessity for models that generate reliable and ethical responses and resonate with professional healthcare practices.

The proposed research contributes to the broader discourse on AI in healthcare by offering a novel framework for assessing AI-generated responses through a blend of empirical user studies and theoretical analysis. This framework aims to mitigate the subjective nature of current evaluations and provide a more robust, standardized approach to assessing the impact of AI technologies on healthcare outcomes. Through this research, we aim to forge a path toward more responsive, responsible, and user-centred AI tools in healthcare, ensuring that they align with both professional standards and patient needs.

In today’s world, products are increasingly intertwined with our daily lives, becoming more intricate. To design these interactions effectively, designers need to comprehend both the people involved in the activities and the technical systems. As user-centered design spreads, the focus shifts from viewing design as problemsolving to recognizing it as the social creation of new possibilities. This shift highlights the importance of users and their everyday lives, as well as their interactions with the product in designing.
Videos serve as a powerful tool for learning about users, their behaviors, and interactions. Video Based Design (VBD) involves analyzing videos to interpret, infer, question actions, and understand the user’s thought process. However, this approach can extend the time and analytical rigor needed due to the extensive analysis required to draw valid conclusions from the video content. To streamline this process and enhance researchers’ workflow, we prototyped a tool that can produce mind maps from videos. These mind maps serve as a visual representation of the video’s key concepts in an organized manner. This tool utilizes Large Language Models (LLMs) to process multimodal video inputs and output the generated mind maps.
To assess the impact of mind maps on the video-based design process, an experimental study was conducted involving 28 designers. Participants were asked to watch two videos from different contexts and then engage in activities using mind maps generated by both the LLM and a human designer. Through quantitative and qualitative analysis of the study data, we gained valuable insights, identified strengths and weaknesses, and proposed design enhancements for the next version of the tool.

Introduction: Individuals with burnout often experience cognitive challenges in general performance or the excessive effort to perform certain tasks. With the increasing work pace over the years, on average, over 10% of workers in Europe and over 17% in non-European countries had experienced burnout up until 2018 (Schaufeli, 2018). Personal solutions like websites and guided meditation applications have emerged to help alleviate burnout symptoms. Governments have passed legislation to prevent burnout from occurring.

The advancing cognitive personal informatics (CPI) technology has made it possible to monitor psychophysiological data and quantify human cognition and cognitive activities. The quantified human cognition-related data could be of great value to people who wish to be more aware of their condition or needs and refer to this information in reflective activities to become a better version of themselves.

Methodology: Three main research questions were generated to inquire about the role of cognitive data in burnout solutions. This project follows the principles of participatory design and the double diamond framework with minor adaptations to answer the research questions. Background research was conducted to gather theoretical and practical information on topics related to burnout solutions, measures of cognitive data, and state-of-the-art CPI technology. Semi-structured interviews with people with burnout experience were conducted as a user study. Creative sessions with target users were conducted to guide the design direction.

Results: The background research clarified the basic burnout profile as well as the cognitive burnout correlates that can be measured by CPI systems. This part of the research elaborated on 7 key correlates that can later be applied in the design phase. The participatory design research activities revealed different perspectives from potential users on burnout that complemented the background research. The participatory design activities also proposed key features that shaped the initial concept.

Conclusion: A design guideline was proposed to help users make good use of CPI technology and self-reflection to search for burnout solutions independently based on the research findings. Some of the key stages in the guideline were conceptualized into wireframes. The guideline suggests users should have full control and support over the entire burnout recovery process, with resources for community support when needed.

Hepatopancreato-biliary (HPB) surgeons, when planning complex cases of pancreatic cancer, lose precious time because of difficulties in visualising the interaction between tumour and blood vessels and identifying patient-specific anatomy. To tackle this issue, the Integrated Imaging Workstation (IIW) was prototyped by e/MTIC and Philips. This prototype encompasses a conventional medical viewer and a holographic display to visualise the same medical data, providing a functional basis and showing promise. However, it lacks a human-centred interaction design, particularly in integrating it into a surgeon’s established interaction flows with conventional medical imaging, as a study with 13 expert surgeons highlights. The literature review and empirical research show that surgeons iteratively scroll through scans to build a mental 3D reconstruction of the anatomy, and have two personal approaches to finding anatomical landmarks and interactions, “vertical scan” and “zoom in and out”. Also, it unveils 3 key design opportunities: minimising workflow disruption, visualising both overview and detail and balancing speed and control in the interaction.

These findings were translated into multiple design concepts, such as one with different modes of visualisation depending on anatomical landmarks. The most desirable concepts were combined into a navigation system, which in turn went through one expert feedback iteration. The final design comprises a system for surgeons to easily select anatomical landmarks through 2D medical imaging. This enables reaching and controlling desired points of view, which then can be orbited around. A high-fidelity prototype was implemented with Unity3D and evaluated. The mixed-method evaluation study with 3 expert HPB surgeons shows promise towards the overall usability of the navigation system and its clinical application. However, some aspects of the interaction, such as the selection of minor vessel branches and missing visual feedback for point-of-view changes, need refinement. These results illustrate a solid human-centred design direction for supporting surgeons in finding crucial landmarks and vascular interactions, as well as an approach to integrate 3D medical visualisation into established workflow with 2D imaging of HPB surgeons.

This thesis hopes to be a starting point for future iterations of the IIW. For example, using metaphors to help the surgeons establish a mental model of the 3D camera system could be tested, or scoping the main user focus to novice surgeons could be helpful to refine the design. Furthermore, this project could be relevant for other research and design endeavours that aim to inject 3D medical visualisation technology into existing workflows in a surgeon-friendly manner.

Dementia is an umbrella term for progressive loss of cognitive abilities, which mainly affects elderly- affecting relationships, feelings and behaviour. A person living with dementia (PLWD) might forget the names of their grandchildren, the face of their loved ones, important places and life events. Due to the rising population growth with more than 55 million people living with dementia (8.1 % of women and 5.4% of men over 65 years) and increasing life expectancies, dementia is already recognized by the WHO (World Health Organization) as a public health priority, and it is now the 7th leading cause of death (WHO, 2021).

The population pressure that care homes are currently under cannot be easily remedied without significant monetary investment into the manpower of the industry. Therefore it would be more fruitful to enable PLWD to function more independently from the help of their carers.

VR technology has seen immense growth in the last 10 years especially within the field of child and elderly care. Although dementia has been understood since 700BCE, the decrease in stigma has allowed the current era of dementia care to flourish using music, art and drama therapy- amongst other person-centred care techniques. Combining these reminiscence therapy techniques with the booming technology of VR is therefore the next logical step in improving immersion. To this end, VR offers control and consistency in testing virtual environments that guarantee safety. However, a key obstacle in the uptake of VR for PLWD is the risk of behavioural and psychological symptoms of dementia- negative behaviours. To reduce these behaviours, a balance must be struck between stimulation and overload- namely stress.

Emotion classification models were established using the valence-arousal circumplex model before identifying the physiological embodiments of such emotion. Of the stress sensing technology surveyed, electromyography was the only technology reviewed that showed evidence of accurate, low-latency, sensation of arousal and valence by way of measuring contractions in the trapezius and zygomaticus muscles respectively.

To design a conducive VR environment, overarching goals for reminiscence in VR were identified as presence, occupation, comfort, inclusion and novelty. The method of reminiscence therapy was broken down into three forms- place, object and interaction based. The parameters of brightness, volume and non-player character movement were established to be adapted in VR for stress reduction over the course of a 10 minute exploration test.

Results of the study showed a 26% decrease in average stress experienced during the three modulation stages over the control stage. Volume modulation was identified as the most effective parameter showing a 32% stress reduction.

Participants reported that the presentation of familiar environments in VR did in fact promote feelings of nostalgia and reported moments of reminiscence. Future work opportunities call for more studies with PLWD, multi-modal & multi-channel stress sensing and concurrent parameter modulation.

Technology improvements such as integrating all peripherals into a standalone system as well as utilising more convenient and ergonomic sensors that could be applied semi-permanently to form a more holistic data overview for PLWD outside experimental conditions.

With an ageing society and fewer people willing to work in healthcare, the pressure on the healthcare sector is growing. And next to that, the high workload increased even more with the COVID-19 pandemic. Next to these factors increase, caregivers need to make numerous decisions each day, is their work emotionally demanding and though to be done. When continuing to work in this way, the quality of care provision will decrease. Therefore technical solutions, like Smart Glasses, can be a way to reduce these high pressures and maintain or even improve the quality of care.

The company 1Minuut already offers an option with Smart Glasses to make knowledge easily, safely and everywhere available to the care provider. Two different Smart Glasses, the VuzixM400 and Google Enterprise 2, and their software Genzõ are already used in several elderly care organisations. Smart Glasses are mainly used to let an expert, like a wound specialist, remotely watch clients from another location. A nurse wears the Smart Glasses, and when video calling with the device, the expert can remotely watch along.

This project focused on two parts. First, the current use and experiences of Smart
Glasses (mainly of 1Minuut) are explored and researched. The second part of the research consisted of developing a design proposal of how Smart Glasses can contribute to supporting caregivers in elderly wound care.

The first part started with explorative activities in the context of Smart Glasses. A better understanding was created of how the Smart Glasses and Genzõ of 1Minuut are
working, are used, and by whom. Then literature research was performed to obtain more information on the experiences of other Smart Glasses users and researchers.
Furthermore, two questionnaires and interviews were conducted by seven participants (clients of 1Minuut) who had all experiences with using Smart Glasses in the actual context of elderly care. A content analysis method was used to analyse the interviews. The project also concentrated partly on what is happening in the elderly (wound) care sector. So when coming up with a proposed design, in the end, this design would contribute to something that can improve or make the work of the elderly care staff easier.

The main results of the study were that Smart Glasses do offer many advantages,
especially for the person watching along and the clients. Less effort needs to be made when experts can see an elderly client remotely, and time will be saved. Nurses receive
expertise on distance and can act faster, which improves the provision of care for the
elderly.

However, Smart Glasses are not yet fully used. This is mainly because there are not yet that many opportunities with and features for the Smart Glasses. In addition, nurses do not yet experience what is really in it for them. Some, especially elderly nurses, find it challenging to control. In conclusion, it takes them more time than it benefits them. Therefore nurses do not use the Smart Glasses much and choose the old manner.

Furthermore, current users mainly use the Smart Glasses when a call is scheduled. They mention that good communication between the nurse and the expert at a distance is essential, as is the reason for use. Therefore an implementation plan is necessary.

The second part of the study focused on proposing a design for how Smart Glasses can support caregivers in elderly wound care. In a short 5-step design sprint, a reporting tool for the Smart Glasses was proposed. This feature helps nurses report wounds with the Smart Glasses by voice. Smart Glasses wearers will be guided through different steps of the report. Closed questions are presented about the wounds, and commonly used answers are shown, which can be selected by saying the answer out loud. There is always an option to add another answer or report at another moment. Reporting can happen now directly in the client’s room after the nurse has finished the tasks. When finished reporting, it will be automatically uploaded to the client’s information file. This makes reporting more accessible, faster and more consistent.

This thesis project aims at improving inventory management and awareness of craftsmen in their light commercial vehicles, specifically in the trunk area. The term commercial vehicle is used to identify a vehicle which main purpose is to transport goods or materials. The light commercial vehicle market, comprising vans and trucks, in Europe is growing rapidly every year and craftsmen are one of the important customers of them. They see their vans as trustworthy colleagues that help them to bring their materials everywhere. However the current state of their vans with different storage solutions do not contribute for solving the issues that they face during intense work days. Those built storage solutions are only focusing on providing space for carried materials and does not contribute to craftsmen’s work performance in any way. In fact this approach reveals major user and context related issues that downgrades the productivity due to the lack of practicality. This status quo is questioned and a smarter solutions to improve inventory awareness and accessibility developed with this graduation thesis. The particular topic is analysed with a human-centric approach. Interviews were conducted with multiple craftsmen to define the problems that they face with, understand their actual needs and observe their van’s trunk area with carried materials. Deficiency in inventory management which is a symptom of several major difficulties faced within the trunk, results as disruptions to the workflow. Having a low-stock of a certain material, travelling additionally to the supplier to cover the deficiency and physically effortful routine controls for the inventory are some of the aspects that causes the disruptions and impracticalities.

The project’s outcome is a product and service design that offers automated control system dedicated for consumables which are small sized materials with high frequency of use. Screws, bolts, fasteners, elbows are some of them that are always carried in the trunk area. The concept utilises accessible sensor technologies integrated to a tangible unit in order to provide momentarily and accurate informations regarding the state of consumables. It enables the connectivity between the inventory, the van and the craftsman by enabling continuous monitoring. The flexibility of the system results as an adaptation to craftsman’s way of work by offering personal settings and configuration in order to be able to procure relevant and suggestive feedbacks, to ultimately improve their inventory management and work efficiency.