Personal Robot
Rethinking Collaboration in the Age of Intelligent Systems
This bachelor’s thesis explored the concept of a social personal robot developed in collaboration with MediaMarktSaturn. The project was written before advanced AI systems became publicly accessible, which makes the concept especially interesting in hindsight: many of the interaction challenges I addressed through design strategies, scripted behaviour, and simulation are becoming increasingly feasible today.
The starting point of the thesis was the observation that social robots have long been part of our cultural imagination, yet very few have become commercially successful. My argument was that this is not only a technological issue, but also a conceptual one.
Many robots are designed as assistants or tools, while their social and emotional role in everyday life remains underdeveloped. In response, I developed the concept of a Personal Robot, or PR, as a new product category comparable to the Personal Computer: a robot that belongs to one specific person and adapts to their identity, routines, preferences, and social life.
I was especially interested in how people perceive robots as social beings, what creates acceptance or discomfort, and how a robot can feel present and meaningful without imitating a human. One important design direction that came out of this research was that the PR should not try to become human. Instead, it should become something different: a believable, expressive, personal companion with its own identity.
The project began with a broad theoretical analysis of social robotics, human-robot interaction, anthropomorphism, personalization, ethics, and the role of robots in society. This research phase formed the foundation of the concept.
The design briefing translated this theoretical foundation into a concrete product challenge.
The goal was to create a physical and tangible communication system between MediaMarktSaturn and the consumer, similar in strategic role to a voice assistant, but more emotional, individual, and embodied. The robot was meant to strengthen the customer relationship by connecting the digital market with the physical, touchable qualities that distinguish MediaMarktSaturn from purely online competitors.
At the same time, the product needed to reflect the company’s customer-centered values. The aesthetic design should be individually adaptable, echoing the idea that MediaMarktSaturn offers fitting solutions for different people, needs, and budgets. More importantly, the concept should take customers seriously not only as consumers with functional needs, but as people with emotions, social relationships, and personal identities.
The resulting design challenge was to create a robot that stands apart from existing social robots by placing individual social needs at the center. It should not replace human interaction, but actively support it: for example by helping people initiate contact, maintain friendships, or create shared experiences. In that sense, the PR is a home assistant on legs, but with character. Its mobility and autonomy are central, because they allow the robot to be perceived less as a device and more as a small creature with its own presence.
Communication
Communication became one of the most important parts of the concept. A robot that is meant to feel social cannot rely only on spoken language or command-based interaction. It needs dialogue, but it also needs nonverbal behaviour: facial expression, gesture, posture, movement, and touch. These elements help communicate emotion and intention, and make the interaction feel more natural and engaging.
For the PR, gesture was especially important because it defines the robot’s physical presence in space. A robot that has done something wrong could make itself small and reduce its presence. A robot seeking attention could become larger, more expressive, and more demanding. These small behavioural shifts allow users to interpret intention, even if the system itself is still technically limited.
Touch was another key interface. Although robots have physical bodies, many are not designed for meaningful physical interaction. In this concept, touch becomes both emotional communication and system feedback. Stroking the robot can act as reward, while a small slap or corrective gesture can signal disapproval. The same logic could extend to voice or expression, such as scolding or giving a stern look. Physical interaction therefore becomes part of training the robot’s behaviour in a way that feels more natural than abstract settings menus.
The robot’s relationship to its owner was deliberately framed with care.
A PR should not simulate or replace human friendship. It should be allowed to become emotionally meaningful, but without pretending to be human. Social robots have the opportunity to become something else entirely: not a person, not just a machine, but a new kind of social object.
For this relationship to develop, the user needs to participate in the robot’s growth. The PR should develop cognitively and character-wise over time, shaped by emotional and physical interaction. Its personality should include not only strengths, but also weaknesses and imperfections. These imperfections are important because they make the robot feel less like a perfect machine and more like a distinct social presence.
Entertainment was therefore not a secondary feature, but a central part of the concept. The PR does not need to be purely practical. It needs to be enjoyable enough that people want to spend time with it. Its behaviour would be built from categorized interaction patterns, such as comments, questions, gestures, animations, and touch reactions. Through trial and error, the robot would learn what kind of behaviour its owner enjoys, accepts, rejects, or ignores.
Back when the thesis was written, this required a more controlled content model based on predefined behaviour patterns and a curated database. Today, with more advanced AI systems, some of these limitations would be much easier to address. Still, the underlying design principle remains relevant: the robot’s behaviour needs boundaries, personality, and a clear relationship to the user’s feedback.
A particularly important part of the concept was the social integration of multiple PRs. The robot should not isolate its owner, but create new reasons for people to meet. PRs could interact with each other, form relationships, and play out small stories over time, almost like a live sitcom between robotic characters. Owners could meet under the pretext of their robots interacting, similar to how dog owners meet through their pets. In this way, the PR becomes a social catalyst rather than a substitute for human connection.
Personalisation
Personalization was another major pillar of the concept. Products that people strongly identify with can become part of their extended self, and this is especially visible in avatar customization in games. The PR applies this logic to a physical object. Users should be able to personalize not only the robot’s appearance, but also its behaviour and personality. The more they invest in shaping it, the more the robot becomes “theirs.”
This personalization also shaped the business concept. The robot would be produced as a standardized base product and then adapted during use. Cosmetic upgrades could be unlocked or purchased through MediaMarktSaturn, creating a connection between personalization, customer loyalty, and store visits. The PR would therefore act as both an emotional companion and a brand touchpoint, individually representing MediaMarktSaturn in the customer’s life.
The PR would also serve functional roles for MediaMarktSaturn.
It could act as a physical key to the digital market, authenticating the user for online shopping or payment through technologies such as fingerprint recognition. Inside the online shop, it could become both avatar and navigation interface. Outside of it, it could function as a personalized shop window, recommending products and offers based on the user’s preferences and behaviour.
Beyond brand-related functions, the PR would also provide classic assistance features such as calendar support, reminders, alarms, notes, and smart home interaction. However, even these functions were imagined through the lens of character. A PR could control smart home devices, but also use them playfully as part of its own “life,” such as secretly flirting with the toaster when it thinks nobody is watching. This kind of detail was important to the concept because it gave the robot more than utility: it gave it personality.
Evaluation
Since building a fully functional social robot was not feasible within the scope of a bachelor’s thesis, I evaluated the concept through an interactive online simulation. The evaluation focused on both utilitarian and hedonic aspects: not only whether people found the idea useful, but also whether they found it enjoyable, engaging, and emotionally acceptable.
Wireframes
I designed the survey as a conversational experience rather than a traditional questionnaire. Participants interacted with a simulated robot through a web interface, moving through an onboarding, dialogue, personalization, and evaluation flow. This allowed me to test parts of the concept in a more experience-based way, even without a physical prototype.
Conversational Elements in the Survey
The conversational interface simulated the robot’s personality and reactions. Users could respond to questions, receive small conversational reactions, and experience the robot less as a static survey object and more as an interaction partner. This was important because the thesis was not only about asking users what they think of a robot, but about letting them experience a small part of the relationship dynamic.
Personalisation in the Survey
The personalization mode allowed users to adapt the robot visually and experience the idea of customization directly. This was one of the central proof points of the concept, because personalization was not just a feature, but a mechanism for identification and attachment. By giving users control over the robot’s appearance and behaviour, the simulation made the idea of a personal robot more tangible.
Onboarding in the Survey
The onboarding introduced the concept step by step. This was necessary because, at the time, many people had little experience with social robots or AI-like interaction. The onboarding therefore had to explain the robot’s role, introduce its behaviour, and make the interaction understandable without overwhelming participants.
Sketching
Design
The final design phase explored different physical forms for the PR. These sketches focused on expressiveness, movement, gesture, and the robot’s ability to support personalization. The goal was not to create a humanoid figure, but a “new character”: something friendly, adaptable, and clearly robotic, avoiding both the expectations and discomfort that often come with human-like robots.
But his operations taken together are so insignificant, a little chipping, baking, patching, and washing, that in an impression so grand as that of the world on the human mind, they do not vary the result.
Looking back, this thesis is an early but very representative example of my design approach. I started from a broad societal and technological question, grounded it in theory, translated it into a product concept, designed an interactive evaluation method, and used storytelling and sketches to make a future-facing idea tangible. The project also shows a design interest that still runs through my work today: how technology can become more meaningful when it supports human needs, relationships, and identity instead of reducing people to users of functions.
