Internet of Things - Final Report

Members of the group:

Name ID Number

Jaskirat Singh Kalra 2021A3PS1020H

Ankita Behera 2021B2A83133H

Arjun Reddy Chinepalli 2021B4A32790H

Satya Guduru 2021B4A82503H

Abhinn Prakhar 2022A3PS0623H

Bodapudi Uday Shankar 2022A3PS0655H

Course details: Internet of Things / EEE F411

Instructor Incharge: Prof Ravi Kiran Y
HD mentor: Sandeep Kumar Maya Nagarajan
Group number: 12
Table of contents

Acknowledgements..................................................................................................................... 3
Smart Attendance Monitoring System Using RFID and AI-Powered Facial Recognition..... 5
Methodology................................................................................................................................ 6
Implementation............................................................................................................................ 7
Comprehensive Comparative Analysis of Solution and the existing systems....................12
Detailed comparison table of the existing system and the solution.................................... 14
Conclusion................................................................................................................................. 15
References................................................................................................................................. 16
Acknowledgements

We would like to express our sincere gratitude to Professor Ravi Kiran Yeleswarapu for his
invaluable guidance, continuous support, and encouragement throughout the course of this
project. His expertise and insights were instrumental in shaping the direction and helped us
make right decisions.

Their constructive feedback, advice, and unwavering patience have been essential in helping us
overcome challenges and refine the project. We deeply appreciate the time and effort our
professor has dedicated to reviewing ideas and providing thoughtful suggestions.

We are profoundly thankful for the opportunities provided, and his mentorship has enriched our
academic and personal growth.

We would like to also express our sincere gratitude to our HD project mentor, Sandeep Kumar
Maya Nagarajan who suggested necessary changes throughout the course and reminded us of
our work.
Summary of the mid semester report

The report outlines the design and implementation of IoT-based technologies for creating smart
campus environments, focusing on Smart Attendance, Smart Energy Savings, and Smart
Security Systems. These solutions aim to enhance efficiency, resource optimization, and safety
on university campuses.

Key Highlights:
Smart Attendance System:

Technology: Uses RFID-enabled ID cards, GPS, GSM modules, and cloud storage for
attendance automation.
Features: Real-time attendance tracking, prevention of proxy attendance, and automated
parental notifications.
Benefits: Reduces errors, saves time, and ensures transparency.
Smart Energy Savings System:

Technology: Combines IoT sensors, smart HVAC systems, and cloud-based analytics to
dynamically adjust energy usage.
Features: Monitors environmental conditions and room occupancy to control heating, cooling,
and ventilation in real-time.
Benefits: Achieves up to 30% energy savings and contributes to sustainability.
Smart Security System:

Technology: Integrates RFID access control, Bluetooth beacons, IoT cameras, and real-time
notification systems.
Features: Monitors campus movements, provides 24/7 surveillance, and issues automated
emergency alerts with evacuation guidance.
Benefits: Enhances campus safety and enables efficient emergency responses.
Conclusion:
IoT technologies offer transformative potential for university campuses by automating
attendance, optimizing energy use, and improving security. While challenges like scalability,
costs, and data security persist, the report highlights the long-term value of these investments
for modernizing campus infrastructure.

However, for the rest of the project, our main focus was implementation of RFID attendance
system in universities.
Smart Attendance Monitoring System Using RFID and AI-Powered
Facial Recognition

Introduction

The process of attendance monitoring in educational institutions is essential for ensuring

student participation and engagement. Traditional methods, such as manual roll calls or
paper-based systems, are prone to errors and inefficiencies. Even more advanced systems,
such as biometric attendance (fingerprint or retina scans), can be costly, slow, and raise privacy
concerns.

In this project, we propose a Smart Attendance Monitoring System that combines RFID
technology with AI-powered facial recognition to automate the attendance process and prevent
attendance fraud. The primary issue addressed by this system is the prevalent proxy
attendance problem, where students use another person’s RFID card to log attendance. By
integrating facial recognition as a secondary layer of verification, this system ensures that only
the registered student can log their attendance, significantly reducing the chances of fraudulent
activities.
Methodology

The methodology behind this system is to use two technologies: RFID and facial recognition, to
verify the identity of the student before logging their attendance. The system consists of several
components:

RFID Technology:

Each student is issued an RFID-enabled card containing a unique ID.

RFID readers are installed at classroom entrances to scan students’ RFID cards when they
arrive.
The scanned RFID ID is sent to the backend for verification.
AI-powered Facial Recognition:

A camera captures the face of the student when they scan their RFID card.
AI-powered facial recognition compares the captured face with the student’s pre-registered
photo in the database.
Attendance is logged only if both the RFID ID and facial recognition match.
Data Verification and Logging:

The system checks whether the student’s RFID ID matches the face captured in real-time.
If both match, the student’s attendance is logged in the database.
If there is a mismatch, the system denies logging attendance and sends an alert.
Parent Notification:

After successful attendance verification, the system sends an SMS notification to the student’s
parents using the GSM module, informing them of their child’s attendance.
 Implementation

The system is implemented in two main parts: the hardware (Arduino) side for RFID scanning
and the software (Python) side for facial recognition.

Hardware: Arduino Setup

The Arduino Uno is used to interface with the RFID reader to capture student RFID card IDs
and send the data to the server for further processing.

Arduino Code (for RFID)

#include <SPI.h>
#include <MFRC522.h>
#include <ESP8266WiFi.h>
#include <WiFiClient.h>

// RFID Configuration
#define RST_PIN 9 // Reset pin for the RFID module
#define SS_PIN 10 // Slave select pin for the RFID module

MFRC522 mfrc522(SS_PIN, RST_PIN); // Create MFRC522 instance

// Wi-Fi Credentials
const char* ssid = "YourWiFiSSID"; // Replace with your Wi-Fi SSID
const char* password = "YourWiFiPassword"; // Replace with your Wi-Fi Password

// Server Details
const char* serverName = "http://yourserver.com/submit_rfid.php"; // Replace with your server
URL

void setup() {
// Initialize Serial Communication
Serial.begin(9600);
SPI.begin(); // Initialize SPI bus for RFID
mfrc522.PCD_Init(); // Initialize RFID reader

// Connect to Wi-Fi
Serial.print("Connecting to Wi-Fi");
WiFi.begin(ssid, password);
while (WiFi.status() != WL_CONNECTED) {
delay(1000);
Serial.print(".");
}
 Serial.println("\nConnected to Wi-Fi!");
}

void loop() {
// Check for a new RFID card
if (mfrc522.PICC_IsNewCardPresent()) {
if (mfrc522.PICC_ReadCardSerial()) {
// Extract RFID card ID
String rfidId = "";
for (byte i = 0; i < mfrc522.uid.size; i++) {
rfidId += String(mfrc522.uid.uidByte[i], HEX);
}
Serial.println("RFID card detected: " + rfidId);

// Send the RFID ID to the server

sendRFIDtoServer(rfidId);
}
}
}

void sendRFIDtoServer(String rfidId) {

if (WiFi.status() == WL_CONNECTED) {
WiFiClient client;

// Prepare POST data

String postData = "rfidId=" + rfidId;

// Connect to the server

if (client.connect(serverName, 80)) {
Serial.println("Connected to server");

// Send HTTP POST Request

client.println("POST /submit_rfid.php HTTP/1.1");
client.println("Host: yourserver.com");
client.println("Content-Type: application/x-www-form-urlencoded");
client.print("Content-Length: ");
client.println(postData.length());
client.println(); // End of headers
client.println(postData); // POST data

// Wait for server response

while (client.connected()) {
if (client.available()) {
String response = client.readString();
 Serial.println("Server Response: " + response);
break;
}
}

client.stop(); // Close connection

Serial.println("Disconnected from server");
} else {
Serial.println("Failed to connect to server");
}
} else {
Serial.println("Wi-Fi not connected");
}
}

Explanation of Arduino Code:

The RFID reader is initialized using the MFRC522 library.
When a student scans their RFID card, the card’s unique ID is captured.
The RFID ID is sent to the backend for comparison with the student’s facial recognition data.
3.2 Software: Python (Facial Recognition)
The facial recognition system is implemented using Python, leveraging OpenCV and dlib for
face detection, and face_recognition for comparing the captured face with the stored face
encoding.

Python Code (for Facial Recognition)

import cv2
import face_recognition

def register_student_face(rfid_id, student_name):

video_capture = cv2.VideoCapture(0) # Use the webcam

print("Capturing face for " + student_name + "...")

# Capture an image
ret, frame = video_capture.read()

# Find face landmarks in the image

rgb_frame = frame[:, :, ::-1] # Convert image to RGB
face_locations = face_recognition.face_locations(rgb_frame)

if face_locations:
# Just take the first face detected
top, right, bottom, left = face_locations[0]
 # Save the student's face encoding and link to their RFID ID
face_encoding = face_recognition.face_encodings(rgb_frame)[0]

# Store the face encoding in the database along with the student's RFID ID
save_face_to_db(rfid_id, face_encoding) # Placeholder function

video_capture.release()

def save_face_to_db(rfid_id, face_encoding):

# Code to store the face encoding in a database (linking it with the RFID ID)
pass

def verify_student_face(rfid_id):
# Get the stored face encoding for the student associated with this RFID
student_face_encoding = fetch_face_encoding_from_db(rfid_id) # Placeholder function

if not student_face_encoding:
print("No face found for this student!")
return False

# Initialize webcam
video_capture = cv2.VideoCapture(0)

while True:
ret, frame = video_capture.read()
rgb_frame = frame[:, :, ::-1] # Convert to RGB

# Find faces in the image

face_locations = face_recognition.face_locations(rgb_frame)

if face_locations:
face_encodings = face_recognition.face_encodings(rgb_frame, face_locations)

# Compare the captured face encoding with the stored encoding

for face_encoding in face_encodings:
matches = face_recognition.compare_faces([student_face_encoding], face_encoding)

if True in matches:
print("Attendance logged for", rfid_id)
return True
else:
print("Face does not match!")
return False
 # Show the video feed for debugging
cv2.imshow("Facial Recognition", frame)

# Stop after 10 seconds if no face is recognized

if cv2.waitKey(1) & 0xFF == ord('q'):
break

video_capture.release()
cv2.destroyAllWindows()

Explanation of Python Code:

Register Student: The register_student_face function captures the student’s face via the
webcam and stores their face encoding linked to their RFID ID.
Verify Student: The verify_student_face function compares the live webcam feed with the stored
face encoding for that RFID ID.
The system will only mark attendance if both RFID and facial recognition match.
Comprehensive Comparative Analysis of Solution and the existing
systems

In modern educational institutions, several advanced methods for attendance monitoring are
employed. While these technologies address various challenges associated with manual
systems, they each have their own limitations:

1. Biometric Systems (Fingerprint and Facial Recognition)

○ How It Works: These systems require students to authenticate their presence
using fingerprints or facial recognition directly at attendance kiosks.
○ Advantages: Biometric systems are accurate and prevent card-swapping fraud,
as they rely on unique physical characteristics.
○ Drawbacks:
■ Slow processing for large groups, leading to bottlenecks at entrances.
■ High costs associated with biometric hardware and maintenance.
■ Hygiene concerns, particularly with fingerprint scanners, as students
physically touch the devices.

2. Barcode/QR Code Systems

○ How It Works: Students scan unique barcodes or QR codes using devices like
smartphones or classroom-installed scanners.
○ Advantages: Cost-effective and easy to implement.
○ Drawbacks:
■ Vulnerable to proxy attendance, as codes can be shared or replicated.
■ Relies heavily on students carrying their devices and ensuring they are
charged and functional.
3. Mobile-based GPS Systems
○ How It Works: Students’ smartphones provide location data to confirm their
presence within predefined geofences (e.g., classrooms).
○ Advantages: Minimal hardware required; real-time monitoring possible.
○ Drawbacks:
■ Students can leave their devices in the classroom to fake attendance.
■ Accuracy issues, especially in densely packed areas or indoor
environments where GPS signal strength is weak.

4. AI-Driven Attendance with Standalone Facial Recognition

○ How It Works: Cameras installed in classrooms detect and identify students
based on pre-stored images.
○ Advantages: Fully automated and non-intrusive for students.
○ Drawbacks:
■ Vulnerable to errors due to lighting conditions, occlusions, or similar facial
features among students.
■ Privacy concerns related to continuous surveillance and biometric data
storage.
Detailed comparison table of the existing system and the solution

Criteria Biometric QR/Barcode Mobile-based AI-driven Our solution

Systems Systems GPS facial
recognition

Fraud High Low (Codes Moderate Moderate High

Prevention (uniqueness) can be (Devices can (prone to (dual-layer
shared) be left) facial errors) verification)

Scalability Expensive for High (cost High High High

large scale effective)

Speed Slow for large Fast Moderate Fast Fast (RFID

groups ensures
quick

Hardware High Low Low Moderate Moderate

costs (Reuses
existing RFID
systems)

Privacy Moderate Low Low High Moderate

concerns

Ease of use Moderate High Moderate High High

Constraints/drawbacks of the proposed solution:

Hardware Dependency:
Requires RFID readers, cameras, and facial recognition-compatible systems,
which may increase initial setup costs compared to manual systems.

Scalability Issues in Dense Environments:

High student density during peak times could lead to processing delays,
particularly if facial recognition encounters errors or retries.
Dependence on Power and Connectivity:
Both RFID and facial recognition systems require a consistent power supply and
stable internet connectivity (for cloud storage or real-time data syncing).

Conclusion
The proposed Smart Attendance Monitoring System demonstrates a highly efficient and
innovative approach to attendance tracking by integrating RFID technology with AI-powered
facial recognition. This dual-layer verification method effectively addresses the limitations of
existing systems, such as proxy attendance and scalability, while maintaining user privacy and
operational efficiency.

Key Advantages:
Fraud Prevention: The combination of RFID and facial recognition ensures accurate and
tamper-proof attendance logging.
Scalability: The system is cost-effective and adaptable to large-scale deployments, utilizing
existing RFID infrastructure.
Speed and Efficiency: RFID provides quick scanning, while facial recognition offers seamless
secondary verification without delaying processes.
Real-time Data Logging: The integration of cloud storage and GSM modules enables immediate
updates and parent notifications.
User-Friendly: Minimal manual intervention and ease of operation make the system suitable for
diverse educational institutions.
Comparative Edge:
Compared to existing systems like biometrics, barcode/QR codes, and standalone AI facial
recognition, the proposed solution offers the highest level of reliability, scalability, and fraud
prevention. It balances privacy concerns while leveraging advanced technologies to optimize
accuracy and speed.

Future Scope:
With potential enhancements such as deeper integration of AI for predictive analytics,
blockchain for secure data storage, and IoT for broader campus applications, this system can
evolve into a comprehensive Smart Campus Management Tool. Its adoption would represent a
significant step forward in modernizing attendance systems and improving overall campus
efficiency.

References

1) [International journal of advances in engineering and management (IJAEM). Volume 3,

issue 6 June 2021, pp: 1307-1313 www.ijaem.net ISSN: 2395-5252]

2) [Sergio Barroso1 · Pablo Bustos1 · Pedro Núñez1

Received: 14 January 2021 / Accepted: 14 December 2021 / Published online: 30 March

2022

© The Author(s) 2022]

3) ZHIQING ZHOU , HENG YU , AND HESHENG SHI

School of Information Engineering, Pingdingshan University, Pingdingshan 467000,

China

Corresponding author: Zhiqing Zhou ([email protected])

4) https://iotprojectsideas.com/iot-rfid-attendance-system-esp32/

5) circuit.io