This research explores how AI-powered “digital representatives” might address Rotterdam’s democratic deficit, where citizen input loses richness as it moves through municipal processes. In Rotterdam, this disconnect manifests in participation disparities (voter turnout ranging from 60% in affluent areas to 21% elsewhere) and declining institutional trust. Through iterative development and evaluation, the study created AI personas capable of maintaining citizen perspectives in policy discussions. These digital representatives integrated speech recognition, Large Language Models, and voice synthesis to participate in deliberations about municipal issues. Key findings demonstrate that while LLMs can consistently represent citizen viewpoints, their authenticity depends entirely on input quality. Representatives derived from rich interview transcripts preserved nuance and influenced policy discussions, while those created from typical municipal documentation proved inadequate. Voice interaction generated stronger engagement than text interfaces, and individual exploration enhanced subsequent group discussions. Evaluation with Rotterdam civil servants revealed both promising capabilities and significant concerns. Digital representatives successfully challenged assumptions about community needs, but participants expressed legitimate concerns about accuracy and the risk of replacing direct citizen engagement with technological simulation. The research’s primary contribution is not the technology itself but revealing critical deficiencies in how municipalities capture and process citizen knowledge. Current documentation practices often fail to retain the experiential context that gives citizen input its value. Addressing Rotterdam’s democratic challenges requires transforming how institutions value citizen perspectives—preserving context rather than pursuing abstraction. Digital representatives cannot solve Rotterdam’s democratic challenges alone but offer a means to sustain citizen perspectives throughout policy processes, especially for voices easily lost in current systems. Their most potent function may be provoking necessary institutional conversations about how citizen knowledge is valued—helping fill, not just simulate presence in, the empty chair. ... Read more

The In-House Creative Studio faces a significant challenge due to an increasing workload among its team of creatives, which includes visual designers, concept creators, copywriters, producers, photographers, and project/event managers. This team may only experience slight expansion, despite the growing demand for campaign and event materials targeted at three main groups: B2B, B2C, and Talent. Consequently, the Creative Studio is forced to be selective, unable to approve all project requests due to the overwhelming volume of work. Ideally, the studio would like to take on more projects. However, the creative team's efficiency is frequently undermined by the considerable time spent managing internal stakeholders, the misalignment resulting from unclear communications, and inefficient workflows. The primary source of these issues was identified as stakeholders providing incomplete or illogical design briefs.

This graduation project explores how Conversational AI might be used in the context of setting up inter-organisational project briefs to help briefing writers improve their design briefings.

Explorative research on the challenges within creative corperate processes and GenAI opportunities clarified that design briefings are a major bottle neck within creative corporate processes

Recognizing this, it becomes evident that to effectively address this bottleneck, stakeholders need assistance in improving the quality of their design breifings prior to the meeting and without external help from the Creative Studio’s briefings reviewers.
This resulted in creating an conversational agent that enabled stakeholders to self-evaluate and improve the quality of their v1 briefings without external help before the first briefing meeting with the Creative Studio.

The final design was evaluated over the span of two sessions with an internal stakeholder of the case company and the Creative Strategist. In the first session, Jelly’s ability to assist the briefing writer was evaluated. In the subsequent session, the feedback comments of Jelly were compared to that of the Creative Strategist.

The results suggest that while Jelly has been effective in certain areas, such as providing detailed feedback that aligns with feedback standards of the briefing reviewers, it requires refinements in user interface language, feedback customisation, and context-specific content generation. The feedback from Jelly, when accurate, led to noticeable improvements in the quality of v1 briefings, aligning with the goals of enabling the stakeholder to self-evaluate their briefings independently.

The findings suggest that Jelly was found to be capable in offering valuable feedback that improved the quality of the briefing.

This thesis concludes with critical reflections to anticipate the future of GenAI in creative corperate processes. The first anticipation includes proposing a new relationship between GenAI and human creatives in the form of co-performance; where one fulfills a role, the other enhances its performance. The second anticipation states that large e-commerce organisation would have terms in their employee contracts that would consider personalised chatbots created by them as intellectual property of the company. In other words, in such organisations, personalised chatbots would likely be developed through a centralised approach.

Ultimately, the design contributes to the field of conversational AI design approaches in the context of corperate creative processes by providing a practical and reproducible example.
... Read more

This report contains a full description of my graduation project for my Master degree in Design for Interaction (Industrial Design Engineering at Delft University of Technology). I did this project in collaboration with the start-up company Sports-f(x), the creators of the NextRound boxing system.

The NextRound boxing system currently consists of an intelligent data-collecting punch bag, paired with a tablet. This system gives users direct feedback on their workout, and enables them to see their improvement in their workout data.

The system has potential to help many different kinds of people to exercise at home. However, there are many barriers withholding people from getting a bag like this into their homes. The main barrier is that positioning a large product like a punch bag (especially one which also needs access to Wi-Fi and electricity) and its accompanying tablet is quite the task.

My graduation project brief was to solve exactly this problem; helping people to position the NextRound system at home. For this purpose, I designed the NextRound Tulip.

To get to this concept product, I undertook a number of design research and exploration activities. To start off, I researched what kind of people might benefit from owning a boxing system as such. From this research I made five personas for potential users.

I used these personas to recruit people for a co-design session, as I wanted to get to know the values of the potential users. These values would afterwards lead to both a proper Design Goal and a list of Requirements & Wishes, which would guide the product development in later stages. The most important value found during the sessions was that people do not wish this big boxing system to be in the way in the house, so other activities done in the house would not be disturbed.

With the findings from this research I set up a brainstorm session with fellow DfI students to try and satisfy the Design Goal. From these sessions I made a Morphological Chart, which led to six ideas. From these six I chose three to develop further into concepts.

I developed the three chosen ideas into three further iterated concepts. Potential users gave their opinions on these concepts through an online survey, which I used, along with the opinions of the company, to choose to develop the NextRound Tulip further.

I made some improvements on this concept and built a full-scale prototype to test and evaluate the concept with potential users. Lastly, I wrote a list of recommendations for future development for a product based on this concept.

The NextRound Tulip is a concept product which should satisfy the needs of users and Sports-f(x). It can be used with the NextRound punch bag, as well as other punch bags. In my opinion, it is a product solution to help people do boxing workouts at home. ... Read more

For many people around the world, water access is a major problem and a daily struggle. Today, up to almost one billion people still lack access to safely managed drinking-water. Especially in sub-Saharan Africa this remains a major issue for many countries. With this project, Quooker wants to help people in these regions gain access to safe drinking-water and help improve the quality of life.

To limit the scope, and make the project more manageable, the focal point is the urban areas of Ethiopia, and in particular Addis-Ababa. In urban areas, the concern surrounding drinking water is the degrading water quality rather than access to water. Improving the water quality is an area in which Quooker can use its expertise.

In Addis-Ababa, people have access to water from a tap point outside their homes. On average, people can get water from these tap points once a week. To bridge the periods of intermittence in the water supply, people store the water in jerrycans at home. During this ‘Water-Journey’ the water quality degrades due to user interactions and poor sanitation. In the end, only 24% of the water consumed by people can be considered safe to drink.

During this project, I developed a product that helps the people in Addis-Ababa treat their water at home by building on and using the expertises of Quooker. The design process focused on providing people in Addis-Ababa with an affordable, reliable and long-lasting product. To achieve this, different water treatment methods were analysed. UV-C was chosen as the preferred method due to its long lifespan and potential to innovate. Because UV-C is a novel technology in the context, this project became a feasibility study for the implementation of UV-C as a water treatment method in Addis-Ababa.

The end-result is the UV-Tap (figure 1), a jerrycan cap with an integrated disinfection system. It disinfects the water with UV-C light at the same time as it dispenses it., which kills the pathogens present in the water. With the UV-Tap a household can treat up to 15.000 L of water before components need to be replaced. This equals to approximately 3.5 years of drinking water for a household of 4 people.
... Read more

Design for product care focusses on stimulating users to prevent early obsolescence of products, which is particularly valuable for electronic household products. One product care strategy is to create attachment and therefore care motivation by strengthening the product-user relationship.
Currently, these strategies lack practicable design directions that designers can use easily, which is why little innovations occur. I propose to stimulate product care behaviour by improving the rela/onship between products and users through ambient interface implementa/on in design.
Ambient interfaces embrace communication with a playful, unobtrusive, intuitive character and gradual presenta/on. The goal of this project was to create a tool that summarises the characteristics of ambient interfaces. This became the ‘Ambient Care-Interaction Index’, an online tool that contains informative text and inspiring examples. It can be used to learn about ambient interfaces and inspire designers to use the qualities to (in)directly stimulate care. ... Read more

Office work is becoming increasingly dynamic, requiring office workers to perform distinct activities rapidly. These activities range from meetings to individual work. Open plan offices (OPO) facilitate workplaces for office workers in a single hall. This causes office workers who perform distinct activities to share the same workspace. These different activities come into conflict with each other, since they produce different sound levels. Furthermore, the open hall of the OPO causes more co-worker interruptions in comparison with offices which are adjusted to occupants performing distinct activities. This thesis collaboration between Advanced Metropolitan Solutions (AMS) and the Living Office Design Lab (LODL) of the faculty of Industrial Design Engineering TU Delft, explores how to facilitate OPO occupants with distinct activities in achieving focus. Through 5 design interventions, a design guideline for achieving office focus is discovered, consisting of 3 elements. 1) Clustering office occupants based on their workstyle preference, decreases co-worker interruptions. 2) For workstyle-based clustering, office occupants need to communicate one of three workstyle preferences: “focus work”, “team work” or “social work”. “Focus work” workstyle communicates that the users don’t want co-worker interruption. “Team work” workstyle communicates that the occupant allows co-worker interruption, as long as the interruption is work related. “Social work” workstyle communicates that the user allows all interruptions, including non-work-related interruptions. 3) Workstyle preferences need to be communicated to the rest of the office by using indicators. Individual indicators are used for desk-to-desk workstyle preference communication. Zone indicators might enable occupants to find workstyle zones within the office. A set of tools that enables OPO occupants to communicate their workstyle preference, through personal and zone indicators, enables workplace selection based on workstyle preference. By enabling occupants to cluster through workstyle-based workplace selection, conflicting activities and co-worker interruptions are reduced. ... Read more

Air pollution is a widely accepted problem in today’s world. Nine out of ten people does not breathe healthy air due to rapid urbanization and human activities that cause increased severe air pollution. Over the years people have had a tendency to comfort themselves with everything they need inside. However, they may not be aware that the amount of dangerous pollutants indoors is often 5 times higher than outside (EPA, 2017). Since we spend 90% of our lifetime indoors, the impact of indoor air pollution (IAP) on our health is relatively high, especially for children. Their bodies and lungs are still developing what makes it very important to breathe safe air. When individuals deal with the risk of being exposed to IAP, they may have various beliefs about the impact it has on their health and the way of keeping indoor air quality (IAQ) levels within an acceptable range. Since people are a valuable asset in buildings, it seems practical to make them able to manage their own air. In the recent years, filtration of air through the use of air cleaners for private use has been an upcoming technological trend. Together with TinyWolf, the startup I am part of, we develop an air purifier to provide various households with clean air where it matters most. For most people air purification introduces new ways of thinking how to improve IAQ. To bridge the research gap of related work in the relationship between IAQ and human behavior, my research aims to understand household members’ perception of air and their related behavior in order to empower and design for their ability to improve the indoor air quality. In order to deal with the complexity of human behavior I used a theory that has found its way from social science into design research. The social practice theory provides a framework to study family and individual life to create an insightful overview of understanding everyday habitual practices that are linked to IAQ. I firstly evaluated the current situation in order to examine the challenges household members face when they want to improve indoor air quality. Findings show the importance of access to information and tools to reduce the exposure to indoor air pollution. Secondly, in the practice-oriented design approach I introduced three households with different versions of TinyWolf’s purifier. The evaluations investigate the influence of the designed products on household members’ perception of the impact of a purifier on the perceived air as well as their ability to adopt the product for use in existing daily practices, in particular the sleeping practice. Findings show that the product alone does not immediately lead to behavior change. It’s meaning and a measure of skill are necessary. Research shows that the empathy between people and product can influence the way they act (Aftab & Rusli, 2007). My project shows a way of how design could be used to empower people’s ability to improve indoor air quality. ... Read more

Agile methods seem to be the answer for today’s turbulent industries, and with it roots in software development, coincides neatly with the increasingly digital world we live in. ABN AMRO also decided to implement a large scale agile framework tailored to their company. Agile methodologies are known to work well at small scale and in an IT context, however at ABN AMRO they’ve implemented it for over 6000 employees from both business and IT departments. Unsurprisingly ABN AMRO faces challenges within this agile organisation. One area in which they’re struggling is managing innovation in a way that aligns with the flow of the tailored and scaled agile organisation. The objective was therefore; To design solution that better aligns innovation processes with the flow of the tailored and scale agile landscape. In addition, literature on large scale tailored agile transformation and innovation management within financial services firms is scarce. Both areas are relevant, however no research was found that combined these topics. Therefor in this thesis we attempt to find challenges for innovation management in a large scale tailored agile framework. In order to both find an answer to this question and help ABN AMRO the challenge was approached from two different streams. Firstly, a revelatory single-case study on ABN AMRO’s tailored agile framework served to understand how they tailored agile at such a large scale and what the challenges were. Secondly, through semi structured interviews with 18 ABN AMRO employees, their innovation management efforts were investigated as well as the challenges they experience with it. Both the challenges from a large scale agile perspective and the company-wide innovation management were merged into a set of six new challenges; “governance on innovation in the agile organisation”, “deficiency of exploration skills”, “getting innovation on the backlog”, “Lack of clear guidance & leadership on innovation”, “Dependencies & Limited Autonomy”, and “Organisational and Cultural legacy”. Based on these challenges a new design challenge was posited which was to design a solution that stimulates ownership and awareness amongst idea owners in the grid landscape towards customer problems and needs. This solution needed to provide clear guidance to enable autonomous and successful realisation of these customer centric innovations in the product and system oriented agile organisation. As a solution to the design challenge and in order to help ABN AMRO with their innovation management efforts, the “User Needs Integration Canvas” was created which is an innovation process that focuses on bridging the gap between customer needs and the complex nature of ABN AMRO’s agile organisation. This study further finds that agile methods don’t cover the innovation process in full, and thus extra efforts and support should go towards managing incremental innovation, the front-end of innovation specifically. In addition traditional financial service providers carry cultural legacy, lack innovation legacy and explorative competences. The role of the individual, with their skills, mindset and behaviour are imperative in in the journey towards a nimbler and more innovative financial service provider. There is no shortcut in becoming an innovative company, whilst a large scale agile framework is a step in the right direction, still a lot of time, effort and patience is needed beyond this step ... Read more