Chatbot for Healthy Ageing
NutriBot is an academic HCI project focused on designing, developing and testing a chatbot that supports healthy ageing through nutrition guidance. Designed for older adults, the prototype provides personalized nutrition advice through a simple conversational interface, using onboarding, user preferences and accessibility settings to create a more supportive user experience.
INDUSTRY
CLIENT
SERVICE
DATE
As people grow older, healthy eating becomes increasingly important, but nutrition advice is not always easy to access, understand or apply in everyday life. Older adults may also have different needs, preferences and levels of digital confidence, which means that a chatbot for this group needs to be simple, trustworthy and accessible.
The challenge was to design a chatbot experience that could provide useful nutrition guidance while still being easy to understand, easy to navigate and adapted to older users.
The goal was to create a high-fidelity chatbot prototype that could help older adults make healthier food choices through personalized and evidence-based nutrition advice.
The project also aimed to explore how older adults experience the usability and accessibility of the interface, how relevant and personalized the chatbot responses feel, and whether users understand the information they receive.
This was a group project where I contributed to the design and development of the NutriBot prototype. The work included research, interface design, prototyping, usability evaluation, user testing and design revision.
The project gave me experience with designing for a specific target group, working with accessibility requirements, evaluating real user feedback and improving an interface through several design iterations.
Scope of Work
Key Features
Main Chatbot Interface
Lets users ask nutrition-related questions in Norwegian through a simple conversational layout.
Personalized Onboarding
Collects allergies, food preferences and health conditions so the chatbot can adapt its responses.
Preference Settings
Allows users to review and update allergies, dietary needs and health-related information after onboarding.
Accessibility Settings
Includes adjustable text size to make the interface easier to read and use for older adults.
Guidance Tutorial
A five-step tutorial added after user testing to explain the chat screen, input field, menu and navigation options.
User Testing & Findings
12
Task-based
Interviews
Onboarding worked, but needed a stronger transition
Users could enter allergies and preferences, but several were unsure what to do when they first reached the main chat screen.
Short questions worked well
Users generally asked short and simple questions and usually received useful answers.
Long responses could feel overwhelming
Some users skimmed longer answers, and weekly meal plans were sometimes experienced as too much text.
Navigation needed clarification
Some users did not realize that menu tabs were clickable, and some had trouble returning to the chat after entering menus.
Personalization was noticed
Users noticed that NutriBot remembered preferences such as allergies, salt preferences and health conditions.
Design Improvements
Guidance Tutorial
Several users were unsure what to do when they first reached the main chat screen.
A five-step tutorial was added to explain the chat screen, input field, menu and navigation options.
Clearer Menu Buttons
Some users did not realize that the menu tabs were clickable.
The menu was redesigned so the tabs looked more like clear interactive buttons.
Back to Chat
Some users had trouble returning to the main chat after entering menus.
A direct “Back to chat” option was added to make navigation easier.
Combined Settings
Settings and accessibility were separated, which made some options harder to find.
Settings and accessibility were combined into one clearer section.
Final Outcome
Reflection
This section reflects on what I learned through the project, the current limitations of the prototype, and opportunities for future improvements.