Student Team
Samridhi Jain
SafeGuard AI Team
Advika Jindal
SafeGuard AI Team
Shaan Garg
SafeGuard AI Team
An AI-powered safety platform that predicts public safety risk levels using location signals, suspicious behaviour indicators, SOS alerts, crowd count, and late-night activity.
Safety risk can rise quickly in isolated or poorly monitored areas, especially during late hours.
Manual monitoring may miss unusual movement patterns, unsafe crowd behaviour, or escalating threat signals.
Different locations carry different safety risk profiles based on crowd density, time, incident history, and alert frequency.
SafeGuard AI combines multiple safety signals into a single risk score so users, responders, and monitoring teams can identify high-risk situations early.
Evaluates area-level safety risk based on location and context.
Flags suspicious behaviour patterns that may indicate danger.
Uses emergency alert frequency and severity as direct safety inputs.
Factors in crowd count and late-night activity to adjust risk levels.
The system receives location, crowd, SOS, suspicious behaviour, and time-of-day data.
AI assigns weights to different safety factors and calculates the current risk level.
The platform classifies areas or incidents as low, medium, high, or critical risk.
Users and responders can receive alerts, prioritize response, and improve situational awareness.
Shows current area risk levels with safety signals and confidence scores.
Ranks SOS alerts so urgent cases can be surfaced first.
Warns users before a location or situation becomes unsafe.
Helps security teams understand where attention is needed most.
Improves over time by learning which combinations of signals often lead to unsafe outcomes.
Designed as a platform for safer campuses, streets, events, and public spaces.