The Smart College Simulator models realistic student and teacher behaviors in a campus environment using Reinforcement Learning (RL) concepts. The system mimics a real-time college schedule where autonomous agents decide to attend classes, cancel lectures, or skip sessions based on internal states and rewards. This simulation powers a live dashboard for tracking class locations, attendance, and faculty status without requiring real-world sensors or manual data entry.
- Simulator: Python (NumPy, Custom RL Agents)
- Backend: Flask (Python)
- Frontend: HTML5, CSS3, Vanilla JavaScript
- Data Storage: JSON (Real-time Simulation State)
- Autonomous Campus: Students and teachers are independent agents that make decisions dynamically, rather than following rigid hard-coded schedules.
- Real-Time Dashboard: A live interface for administrators to monitor ongoing classes and attendance percentages.
- AI Admin Chatbot: Natural language interface for administrators to query specific student statuses or overall campus health.
- Faculty Tracking: Real-time location tracking for teachers based on their current teaching status.
The simulation uses probabilistic models and reward-based logic to drive agent behavior:
- Action Space: Teach, Cancel.
- Logic: Decisions are influenced by assigned time slots and a history of previous cancellations.
- Reward: +10 for teaching, -5 for cancelling.
- Action Space: Attend, Skip.
- Logic: Decisions are based on their current attendance percentage and the status of the class (Ongoing vs Cancelled).
- Reward: +5 for attending. A significant penalty (-20) is applied if attendance falls below 75%, simulating the motivation to avoid detention.
- Install Dependencies:
pip install -r requirements.txt
-
Install Dependencies:
pip install -r requirements.txt
-
Start Application:
python backend/main.py
Note: The Simulator runs automatically in the background.
-
Access Dashboard: Open your browser and navigate to:
http://localhost:8000