AI Assisted Hybrid Indoor and Outdoor Campus Navigator with Privacy Preserving Geofenced Attendance System

The rapid growth of large, multi-building campuses has made navigation difficult for students, staff, and visitors, while existing attendance systems raise privacy concerns from continuous location tracking. Prior navigation systems rely mainly on outdoor GPS, offer little indoor
wayfinding, and use inefficient algorithms like Dijkstra's that can't adapt to real-time changes. Prior attendance systems depend on continuous GPS surveillance and centralized biometric storage, creating battery, privacy, and security risks. The proposed project provides a unified
website-based platform addressing both problems together to investigate navigation and
attendance while reducing unnecessary location tracking. An efficient hybrid indoor-outdoor
navigation system has been designed to develop a privacy-preserving geofence-based attendance
mechanism. An easily accessible and offline-capable solution has been built with measurable
benchmarks. Using Global Positioning System (GPS) for outdoor pathfinding and QR-code or
Bluetooth Low Energy (BLE) checkpoints for indoor localization. The algorithms have been
used for computing the routes that adapt to blocked paths or crowd density. The built-in features
include wheelchair-friendly routes and locally cached offline maps. Attendance uses
event-triggered geofenced check-ins instead of continuous tracking, with an optional
Convolutional Neural Network (CNN)-based facial-verification step storing encrypted
embeddings. Built with a Python backend (Flask/Django) and a responsive website front end.
Lower battery/privacy risk, better indoor wayfinding, offline usability, accessibility, and
encrypted data storage — evaluated via route computation time, positioning accuracy, and
check-in reliability. Indoor accuracy depends on checkpoint density, facial verification needs
good lighting, and the system is untested on a live, large-scale deployment.