Novasci

Improving science learning using AI and 3D models

Comment

Inspiration

We were inspired by the challenge of making abstract science concepts tangible and accessible. Traditional learning can feel passive and flat. We wanted to create a space where students don't just hear about science, but can experience and explore it. Our vision was to build a personalized, interactive classroom that feels less like studying and more like discovery, powered by the latest advances in generative AI.

What it does

Novasci is an immersive, AI-generated 3D classroom environment that delivers personalized science lessons through voice interaction. Users can speak to their AI tutor, which responds conversationally using Google's text-to-speech to explain complex topics. To reinforce learning, Novasci features quick, AI-generated quizzes for brainstorming and active recall, and it tracks learning progress through a simple dashboard. The 3D environment visually stimulates learning, turning subjects like biology and physics into interactive experiences.

How we built it

  • Core AI & Environment: We built the project using Google AI Studio, leveraging Gemini's multimodal capabilities to generate the descriptive framework and educational content for the 3D classroom.
  • Interaction & Voice: We integrated Google's Text-to-Speech (TTS) API to give the AI tutor a natural, human-like voice, enabling the "speak-to-teach" functionality.
  • Quizzes & Logic: The application logic, coded in Python, manages user input, the conversation flow with Gemini, and the generation of quick-review quizzes based on lesson content.
  • Progress Tracking: A lightweight database stores user quiz results and interaction history to power the learning progress tracker.

Challenges we ran into

  • Synchronizing Modalities: Aligning the generative 3D environment descriptions with the sequential, logical flow of a science lesson was a complex task.
  • Latency & Flow: Minimizing the delay between a user's spoken question, the AI's processing, and the audio response to maintain a natural conversational feel.
  • Designing Effective AI Quizzes: Creating a system where Gemini could generate meaningful, topic-relevant quizzes that test understanding rather than just recall, and integrating scoring into the progress tracker.

Accomplishments that we're proud of

  • Creating a seamless, end-to-end experience where a user's voice initiates a dynamic lesson, interactive quizzes, and visual stimulation.
  • Successfully integrating multiple Google AI APIs (Gemini and TTS) into a single, cohesive application that feels greater than the sum of its parts.
  • Building core engagement features—quick quizzes and progress tracking—on top of the immersive environment, moving beyond a demo to a tool with measurable learning outcomes.

What we learned

  • The immense potential and current challenges of using generative AI for structured educational content.
  • The importance of prompt engineering to act as a "curriculum guide" for a powerful but general-purpose LLM like Gemini.
  • How voice interaction and immediate feedback (quizzes) fundamentally increase user engagement and knowledge retention.
  • The technical considerations in designing a simple yet effective progress-tracking system alongside an AI-native application.

What's next for Novasci

  • Interactive 3D Models: Evolving from a described environment to a fully rendered, interactive 3D classroom. Users will manipulate models of cells, planets, or chemical compounds.
  • Enhanced Assessment: Developing more diverse quiz formats (e.g., drag-and-drop, scenario-based questions) and a more detailed analytics dashboard for deeper insight into learning patterns.
  • Multimodal Input: Adding image and object recognition so a user can point their camera at the world around them (e.g., a leaf, a rock) and have Novasci launch an immediate, relevant lesson.
  • Adaptive Learning Paths: Evolving the progress tracker into a personalized learning engine. The AI tutor will adapt lesson difficulty and content in real-time based on quiz performance and user goals.
  • Subject & Language Expansion: Broadening the curriculum to cover the full STEM spectrum and beyond, with plans to make Novasci accessible in multiple languages to learners worldwide.

Built With

Share this project:

Updates