Student Attendance App Project with Bluetooth Beacon Auto-Marking and Offline Roll Number List

College attendance still runs on shouted roll calls and paper registers: a 60-student class can lose 8–10 minutes every period, proxy attendance is trivial to pull off, and the register itself is one fragile copy. The Student Attendance App with Bluetooth Beacon Auto-Marking and Offline Roll Number List replaces the roll call with radio proximity. The teacher's phone broadcasts a Bluetooth Low Energy beacon carrying a unique session ID; every student's phone runs the AttendBeacon app, which scans for that beacon in the background. When the received signal strength (RSSI) stays above a −70 dBm threshold — roughly inside the classroom — the app marks that student's roll number present in the on-device database, with no taps and no internet. The full class roll-number list is stored offline in a Room database, so a whole session runs on zero connectivity; teachers get manual override, a live session dashboard, and date-wise CSV export of attendance.

End to end: the teacher opens the app, selects the class and starts a session — and the teacher's phone begins advertising a Bluetooth Low Energy beacon packet carrying a 16-byte session UUID plus major and minor codes that identify the class and period, rebroadcast roughly every 100 ms. Each student's phone runs a background BLE scan; every received packet carries a signal-strength reading (RSSI) that falls as distance grows. The app keeps only packets matching the session UUID, converts RSSI to an estimated distance with the log-distance formula, and applies the proximity rule: an RSSI of −70 dBm or stronger — about 3 metres — means the student is inside the classroom. The moment that threshold is crossed, the app writes a timestamped PRESENT row for the student's roll number into the on-device Room database, which is also why the whole session runs offline. Weak or unknown beacons are ignored, and the teacher can override any entry manually.

1. The teacher selects the class and starts a session; the teacher's phone advertises a BLE beacon packet — 16-byte session UUID, 2-byte major and 2-byte minor codes — rebroadcast approximately every 100 ms.
2. Each student's phone runs a background BLE scan (BluetoothLeScanner); every received advertisement reports its RSSI, which weakens with distance.
3. The app discards packets whose UUID does not match the session, then converts the beacon's RSSI to an estimated distance using the log-distance formula d = 10^((TxPower − RSSI) / (10 · n)), with TxPower −59 dBm calibrated at 1 m.
4. When the beacon's RSSI reaches −70 dBm or stronger — roughly inside the classroom — the proximity rule fires and the app writes a timestamped PRESENT row for the student's roll number into the on-device Room database.
5. The teacher watches the live session dashboard as rows flip to present; any entry can be corrected by hand and is tagged MANUAL for the record.
6. When the session ends, the teacher exports the date-wise CSV. Everything — roll list, marks, log — stayed on the device; no internet was needed at any point.

Tech stack:

• Kotlin
• Android Jetpack (ViewModel · LiveData · Navigation)
• BluetoothLeScanner · BluetoothLeAdvertiser (Android BLE APIs, API 21+)
• Room database (offline-first storage)
• Kotlin Coroutines · Flow
• Material 3 (Material You UI)
• WorkManager (background scan scheduling)
• Date-wise CSV export

The AttendBeacon app's four most important capabilities are beacon auto-marking, the offline roll-number list, teacher broadcast mode, and the proximity threshold control. Beacon auto-marking runs a continuous BLE scan in student mode; the moment the class beacon's RSSI crosses the −70 dBm threshold, the app writes a timestamped PRESENT row to the local Room database — in the concept build's browser demo, the detect-to-mark UI update measured about 44 ms. The offline roll-number list holds the whole class in an on-device Room database (32 students in the demo seed, expandable to thousands), so sessions work with zero connectivity. Teacher broadcast mode turns the teacher's phone into the classroom beacon, advertising the session UUID with major and minor codes per class and period. The proximity threshold control makes the −70 dBm cutoff configurable per room, so corridor signals and the next classroom's beacon are ignored instead of marking the wrong student.

• [BLE beacon auto-marking] In student mode the app scans continuously for the class session beacon; when its RSSI crosses the −70 dBm proximity threshold, the student's roll number is marked present with a timestamp — no taps, no QR scan, no internet.
• [Offline roll-number list] The entire class list (roll numbers, names, divisions) is stored in an on-device Room database. Teachers can add, search and import students from CSV; a session runs with zero connectivity.
• [Teacher broadcast mode] The teacher's phone advertises the session beacon itself — a 16-byte session UUID with major and minor codes identifying the class and period — so no extra beacon hardware is needed for the demo classroom.
• [Configurable proximity threshold] The −70 dBm cutoff is adjustable per room; weaker signals (corridor, adjacent classroom) and unknown beacons are filtered out before they can mark anyone.
• [Manual override with audit trail] The teacher can mark any student present or absent by hand; manual entries are tagged MANUAL in the list, so beacon-marked and hand-marked rows stay distinguishable.
• [Date-wise CSV export] One tap exports the session's attendance — roll number, name, status, timestamp — as a CSV file the teacher can submit as the official record.
• [Scan log and signal view] Every detection is logged with beacon name, RSSI, estimated distance and the measured detect-to-mark latency, which makes the proximity mechanism demonstrable in a viva.
• [Cloud sync (optional)] Syncing attendance to a department portal or Firebase backend can be added as an optional customization; the base build is fully offline by design.

• Complete Android Studio project source (Kotlin, teacher + student modes)
• Seeded demo dataset: 32-student roll list and session fixtures used in the concept build
• Proximity-marking module with the RSSI threshold and filtering logic, documented for the viva
• Project report PDF (BLE background, proximity mechanism, data model, methodology)
• PPT presentation for final review
• Viva Q&A preparation document (BLE advertising/scanning, RSSI proximity, Room, background execution, permissions)