This white paper presents NudgeMe, a mobile health prototype developed within the Master Programme Digital Healthcare. The prototype explores how large language model (LLM)-based technologies, nutritional science, and behavioral nudging can be combined to support healthier eating habits in everyday life. In addition to describing the existing prototype, the paper outlines conceptual future perspectives for extending NudgeMe through wearable-integrated, context-sensitive just-in-time nudging.
The project is motivated by the rising prevalence of overweight, obesity, and metabolic diseases, as well as the limited long-term effectiveness of conventional nutritio interventions. Although nutritional guidelines are widely available, many individuals struggle to translate this knowledge into sustainable daily behavior due to time pressure, cognitive overload, and the effort required by calorie-based tracking approaches.
The NudgeMe Prototype introduces an LLM-based Meal Plate Assessment grounded in the Plate Model. Users photograph their meals, which are analyzed using multimodal large language models to provide visual and qualitative feedback on plate composition. By focusing on meal balance rather than calorie counting, the prototype aims to reduce cognitive burden and lower barriers to engagement.
From a behavioral perspective, the Prototype applies nudging principles, offering supportive and positively framed feedback instead of directive reminders. Customizable notifications and visual progress indicators are designed to preserve user autonomy and minimize app fatigue.
A key contribution to this white paper is the conceptual extension of the prototype through wearable integration. While not yet implemented, this approach explores how wearable-derived contextual information—such as eating speed, meal duration, or stress-related indicators—could enable just-in-time nudging at moments of decision-making. This multimodal concept aims to complement meal composition analysis with behavioral insights while remaining non-intrusive.
In summary, this white paper positions the NudgeMe Prototype as a functional proof-of-concept, demonstrates the feasibility of LLM-based meal assessment combined with behavioral nudging, and provides a foundation for future research into wearable-integrated, context-sensitive nutrition interventions.
Team Members: Bettina Knabl, Angela Löscher, Niklas Schäfer, Valerie Pramer
Project Coach: Jakob Doppler
A project at the St. Pölten University of Applied Sciences. Master Program Digital Heathcare
