Inspiration
Starting college can be challenging for students who are unfamiliar with the campus layout, academic facilities, and available learning resources. Traditional campus maps help students find buildings, but they do not explain how these facilities support learning or how students can use them to develop practical skills.
We wanted to transform campus navigation into something more meaningful: an interactive learning experience where students can discover educational resources, understand their purpose, and apply their knowledge through practical activities.
This inspired CampusQuest AI, an AI-powered campus learning companion that combines intelligent navigation, academic resource discovery, and personalized learning challenges.
What it does
CampusQuest AI transforms a traditional campus map into an interactive educational platform.
AI Campus Navigator: Understands natural-language queries and helps students locate classrooms, laboratories, libraries, and other campus facilities.
Smart Resource Discovery: Recommends relevant academic resources based on students' interests and learning goals.
Interactive Learning Quests: Generates practical challenges connected to academic subjects, laboratory equipment, and campus facilities.
AI Learning Assistant: Explains concepts in simple language and provides contextual answers using verified campus information.
Personalized Feedback: Evaluates student responses and suggests follow-up activities to improve understanding.
Intelligent Pathfinding: Calculates suitable campus routes using mapped pathways and shortest-path algorithms.
For example, a student learning computer networking can ask for the networking laboratory, receive directions, explore the laboratory's educational resources, and complete a related practical challenge.
How we built it
We designed CampusQuest AI as a web application with an interactive frontend, backend services, a campus knowledge database, and AI-powered educational features.
Frontend: React and Vite provide the user interface, including the campus map, navigation search, learning activities, and student dashboard.
Backend: Node.js and Express handle user requests, retrieve campus information, coordinate AI services, and process application data.
Database: MySQL stores campus buildings, facilities, pathways, educational resources, and student learning progress.
AI and machine learning components:
NLP intent classification: TF-IDF and Logistic Regression can classify student queries into categories such as navigation, facility information, and learning-resource discovery.
Large Language Model (LLM): A model accessed through an API, such as Gemini, generates explanations, practical challenges, and feedback based on the student's question and relevant campus information.
Retrieval-Augmented Generation (RAG): Retrieves relevant information from the campus knowledge base to ground AI-generated responses in verified data.
Content-based recommendation: TF-IDF and cosine similarity can match learning goals with relevant campus resources and activities.
Pathfinding: A* or Dijkstra's algorithm calculates routes across the campus pathway graph.
These components work together to understand student requests, retrieve appropriate information, recommend learning opportunities, and calculate routes. The AI handles language and educational assistance, while the pathfinding algorithm handles route computation.
Challenges we ran into
One of our main challenges was designing a system that combines navigation with meaningful educational assistance rather than simply adding a chatbot to a map.
Another challenge was ensuring that AI-generated responses remain relevant to actual campus facilities. Language models can produce incorrect information, so campus-specific answers need to be grounded in verified records.
Integrating natural-language queries with structured campus data also requires careful handling of different ways students may ask for the same destination or resource.
We plan to address these challenges through structured campus data, retrieval-grounded responses, carefully designed prompts, validated pathways, and testing with realistic student queries.
Accomplishments that we're proud of
We are proud of developing a project concept that connects AI, education, and campus navigation into one practical experience.
Our key accomplishments and goals include:
Designing a unified workflow connecting student queries, campus information, navigation, and educational activities.
Identifying specific AI models and algorithms for different tasks instead of relying on a single general-purpose chatbot.
Creating an approach that connects academic learning goals with real campus resources.
Prioritizing personalized, application-based learning over simple memorization.
Developing a modular architecture that can be extended to different campuses and academic environments.
Our biggest achievement is the idea of transforming campus exploration into an opportunity for continuous learning.
What we learned
Through designing CampusQuest AI, we explored how different AI techniques can work together to solve a real-world problem.
We learned that NLP can help interpret student requests, large language models can make educational assistance more conversational, and retrieval-based methods can improve the relevance of generated responses.
We also recognized that AI is not the right tool for every task. Algorithms such as A* and Dijkstra's are more suitable for calculating routes, while structured databases provide a reliable source of campus information.
Most importantly, we learned that an effective AI application should combine intelligent assistance with reliable data, a clear user experience, and measurable value for its users.
What's next for
Our next goal is to develop and test a functional prototype using real campus data.
Future improvements include:
Interactive campus maps: Support more buildings, pathways, and accessible routes.
Multilingual assistance: Help students interact with the system in English and additional local languages.
Advanced personalization: Recommend learning activities based on student interests, performance, and progress.
Quest-based learning: Introduce achievements, progress tracking, and subject-specific challenges.
Accessibility features: Improve usability for students with different mobility and navigation needs.
Campus integrations: Connect verified facility information, laboratory schedules, and library resources where authorized.
Performance evaluation: Test route accuracy, response relevance, recommendation quality, and student learning outcomes.
Our long-term vision is to make every campus a more accessible and interactive learning environment, where students can find the right place, discover the right resources, and turn everyday campus exploration into meaningful educational experiences.
Built With
- ai
- artificial-intelligence
- education
- express.js
- large-language-models
- machine-learning
- mysql
- natural-language-processing
- node.js
- pathfinding
- python
- react
- recommendation-system
- retrieval-augmented-generation
- scikit-learn
- vite
Log in or sign up for Devpost to join the conversation.