HostelOps – Smart Hostel Resource Management

Inspiration

Hostel management is often reactive and fragmented. Electricity, water, and Wi-Fi usage are monitored manually, while maintenance issues are handled only after problems occur. This makes it difficult to identify unusual consumption, predict future demand, and respond quickly. We wanted to build a system that could answer not only “What is happening?”, but also “Why is it happening?” and “What should be done?”. This inspired us to create HostelOps, an AI-powered platform for predictive hostel resource management.

What it does

HostelOps brings hostel utility monitoring, prediction, anomaly detection, and maintenance into one platform. It:

• Predicts electricity, water, and Wi-Fi demand.
• Monitors room- and floor-level resource usage.
• Detects unusual consumption using AI.
• Explains detected anomalies and suggests actions.
• Automatically converts detected issues into maintenance tickets.
• Ranks rooms and floors based on efficiency.
• Provides role-based access for administrators, wardens, and staff.
• Sends in-app and email notifications. ##How we built it We simulated realistic daily utility usage based on occupancy and academic schedules. We then engineered features such as day of the week, occupancy ratio, lag values, and rolling averages. For prediction, we trained separate Random Forest models for electricity, water, and Wi-Fi. Isolation Forest models are used to detect abnormal consumption and classify anomalies by severity. The backend is built using Flask REST API, while SQLite stores room records, utility readings, occupancy logs, alerts, and maintenance tickets. Firebase Authentication provides role-based access, and HTML, CSS, and JavaScript power the dashboard. The application is deployed using Render.
  ##Challenges we ran into One of our biggest challenges was bringing multiple utilities into a single intelligent system instead of treating electricity, water, and Wi-Fi independently. We also had to design meaningful features from usage and occupancy data, detect genuine anomalies, and convert technical predictions into understandable explanations. Another challenge was creating a complete workflow beyond a read-only dashboard. We wanted a detected anomaly to lead to an actual maintenance ticket and resolution process, rather than simply displaying an alert. ##Accomplishments that we're proud of We are proud of building a unified system that combines forecasting, anomaly detection, explanation, alerts, and maintenance management. Our key achievement is connecting the complete workflow: Utility Data → AI Prediction → Anomaly Detection → Explanation → Maintenance Ticket → Resolution We also implemented room/floor-level monitoring and role-based views, making the platform useful for different hostel management roles. ##What we learned Through HostelOps, we learned how to apply machine learning to a practical real-world problem rather than using AI only for prediction. We gained experience in:
• Machine-learning forecasting
• Anomaly detection
• Feature engineering
• REST API development
• Database management
• Authentication and role-based access
• Dashboard development
• Connecting AI predictions with real operational workflows We also learned that an effective AI system should provide actionable insights, not just predictions. What's next for HostelOps – Smart Hostel Resource Management Our next step is to move from simulated data toward real-time IoT sensor integration for electricity and water meters. We also plan to improve forecasting accuracy with more real-world data, expand the system across multiple hostels and campus buildings, and strengthen sustainability analytics. Ultimately, we want HostelOps to evolve into a campus-wide intelligent resource management platform that predicts problems before they occur, reduces resource wastage, and enables faster maintenance decisions.

Built With

• api
• authentication
• css
• firebase
• flask
• forest
• html
• isolation
• javascript
• python
• random
• rest
• scikit-learn
• smtp
• sqlite