WattWise Pulse

InspirationSmall campuses and community facilities often have devices that waste energy or show early signs of failure, but operators lack a simple monitoring view. We wanted to make the first response understandable instead of burying people in raw sensor data.## What it doesWattWise Pulse turns simulated IoT readings for power, temperature, and vibration into a compact dashboard. It compares readings with a baseline, flags unusual combinations, and explains a practical next action. The prototype is intentionally a monitoring aid, not an automated safety control system.## How we built itThe prototype is a Streamlit app in Python. A reproducible telemetry simulator generates device readings, pandas manages the time series, and a normalized deviation score combines the signals into an explainable anomaly score. Controls let a user choose a device, change the time horizon, inject a test spike, and generate fresh telemetry.## ChallengesThe hardest part was designing an alert that is useful without creating noise. We chose a small end-to-end vertical slice and made the reason for every alert visible. The current simulator makes the demo deterministic and easy to test while keeping the next production steps clear.## What we learnedA good operational AI experience should show both the signal and the reason for the recommendation. Clear thresholds, a visible baseline, and a human decision point matter as much as the model.## What's nextThe next version would add authenticated MQTT ingestion, persistent storage, access controls, historical trend views, tested device-specific thresholds, and notifications that require a human acknowledgement before escalation.## Built for this hackathonThis project is a new, event-specific prototype for VoltHacks. The current version uses simulated telemetry and does not claim to control physical devices.

Built With

• iot-telemetry-simulation
• pandas
• python
• scikit-learn
• streamlit