Tab 1
AI PROJECT LOGBOOK
(TODO: temporary cover, will be
changed for better one, later on)
AI Project Logbook
Project Details:
PROJECT NAME: AI Enabled attendance system
SCHOOL NAME: Techno India Group Public School
YEAR/CLASS: 2025-2026 , 12th
TEACHER NAME: Abhro Banerjee
TEACHER EMAIL:
Team Members
Name Class
Aditi Shukla XII
Astha Agarwal XII
Hrichik Mazumder XII
Ishanika Mukherjee XII
Ishwani Banerjee XII
Prince Singh XII
Simran Singh XII
1. Introduction
This project logbook records the development of an AI Enabled Attendance System.
Marking attendance manually takes valuable class time and is vulnerable to errors and
impersonation.
Our AI-Enabled Attendance System uses advanced facial recognition to automatically
identify students and update the register in real time.
It saves valuable class time each day, ensures higher accuracy, and securely stores records
for instant reporting and analysis. The system makes attendance faster, contactless, and
highly reliable, enhancing both efficiency and convenience.
2. Team Roles
2.1 Members and their roles
Role Role description Team Member Name
Schedules the task among
the team member, Ensures
Project leader the task is completed on Hrichik Mazumder
time, Resolves doubts (if
any), and one source of
contact
Decides upon the data
required, type of data for
Data expert training the model, collects Ishanika Mukherjee
the data, ensures the type of
data, and its authenticity
Collects questions from the
end users which needs to
be answered, look for the
Information researcher answers for those questions Ishwani Banerjee
and then prepare a report
which will be passed to the
project leader
Will create the design and
the flow of how to go about Hrichik Mazumder
Designer Aditi Shukla
making the solution for the
problem statement.
Works to build the model,
Prototype builder/coder & train it followed by testing Hrichik Mazumder
Tester the efficiency and accuracy
of the model
Prepares reports,
maintains project
Documentation & documentation, and
Simran Singh
Presentation Specialist designs presentation Prince Singh
materials.
Acts as the link between
the team and external Astha Agarwal
Communication & stakeholders (teachers, Ishwani Banerjee
Outreach Coordinator judges, users).
Ensures responsible use of
AI, checks for bias, data
privacy, and ethical Ishanika Mukherjee
Ethics Analyst compliance.
2.2 Project plan
Planned Actual
duration duration
Planned Planned (hours, Actual Actual end (hours, Who is
Phase Task start date end date minutes) start date date minutes) responsible Notes/Remarks
Read papers on facial
Coursework, Team recognition & ML
readings 16.09.2025 20.09.2025 12 Hours 15.09.2025 19.09.2025 11.5 Hours Members ethics
Set up a team Folder on Google
Preparing for folder on a Team Drive with shared
the project shared drive 16.09.2025 17.09.2025 2 Hours 16.09.2025 16.09.2025 1.5 Hours Members permissions
Background Team Identified challenges
reading 17.09.2025 18.09.2025 1 Hour 17.09.2025 17.09.2025 1 Hour Members in manual attendance
Research issues Found inefficiencies
in our Team in current attendance
community 18.09.2025 18.09.2025 1.5 Hours 18.09.2025 18.09.2025 1.5 Hours Members systems
Team meeting
to discuss issues
& select project Team Decided on AI-based
idea 18.09.2025 18.09.2025 1 Hour 18.09.2025 18.09.2025 1 Hour Members attendance system
Complete
Defining the section 3 of Team
problem Logbook 19.09.2025 19.09.2025 20 Mins 19.09.2025 19.09.2025 20 Mins Members Collaborative entry
Teachers and
Team students identified as
Identify users 20.09.2025 20.09.2025 30 Mins 20.09.2025 20.09.2025 30 Mins Members users
Meeting with
users to observe Team Observed class
them 21.09.2025 21.09.2025 45 Mins 21.09.2025 21.09.2025 45 Mins Members attendance flow
Interview with Team Teacher feedback
user (1) 21.09.2025 21.09.2025 20 Mins 21.09.2025 21.09.2025 20 Mins Members collected
Interview with Team Student feedback
user (2) 22.09.2025 22.09.2025 15 Mins 22.09.2025 22.09.2025 15 Mins Members collected
Complete
Understanding section 4 of Team
the users Logbook 22.09.2025 22.09.2025 30 Mins 22.09.2025 22.09.2025 30 Mins Members Collaborative work
Team meeting
to generate Team Discussed AI model
ideas 23.09.2025 23.09.2025 1.5 Hours 23.09.2025 23.09.2025 1.5 Hours Members options
Complete
section 5 of Team
Brainstorming Logbook 23.09.2025 24.09.2025 30 Mins 23.09.2025 23.09.2025 30 Mins Members Collaborative work
Team Finalized architecture
Design meeting 24.09.2025 25.09.2025 3 Hours 24.09.2025 24.09.2025 3 Hours Members & modular layout
Complete
Designing the section 6 of Team
solution Logbook 25.09.2025 25.09.2025 30 Mins 25.09.2025 25.09.2025 30 Mins Members Collaborative work
Meeting on data Team Discussed dataset
requirements 25.09.2025 25.09.2025 30 Mins 25.09.2025 25.09.2025 30 Mins Members structure
Collecting &
preparing Team Captured sample face
data Data collection 26.09.2025 27.09.2025 1.5 Hours 26.09.2025 27.09.2025 1 Hour Members images
Planned Actual
duration duration
Planned Planned (hours, Actual Actual end (hours, Who is
Phase Task start date end date minutes) start date date minutes) responsible Notes/Remarks
Data
preparation & Team Cleaned and labeled
labelling 27.09.2025 28.09.2025 2 Hours 27.09.2025 28.09.2025 2 Hours Members dataset
Plan prototyping Team Planned iteration
phase 29.09.2025 29.09.2025 2 Hours 29.09.2025 29.09.2025 2 Hours Members stages
Researched about
Research Team Facial recognition
Solutions 30.09.2025 01.10.2025 4 Hours 30.09.2025 01.10.2025 4 Hours Members solutions
Team
Test Solutions 01.10.2025 02.10.2025 2 Hours 01.10.2025 02.10.2025 2 Hours Members Tested CompreFace
Integrated with
Write program Team Mediapipe &
Prototyping logic 03.10.2025 03.10.2025 45 Mins 03.10.2025 03.10.2025 45 Mins Members CompreFace
Invite users to Team Tested with
test prototype 04.10.2025 04.10.2025 20 Mins 04.10.2025 04.10.2025 20 Mins Members classmates
Conduct testing Team
with users 04.10.2025 04.10.2025 30 Mins 04.10.2025 04.10.2025 30 Mins Members Collected feedback
Logbook Section Team
Testing 9 05.10.2025 05.10.2025 15 Mins 05.10.2025 05.10.2025 15 Mins Members Documented testing
Discuss video Team
plan 06.10.2025 06.10.2025 10 Mins 06.10.2025 06.10.2025 10 Mins Members Planned structure
Team
Write script 07.10.2025 07.10.2025 20 Mins 07.10.2025 07.10.2025 20 Mins Members Drafted narration
Team
Film video 07.10.2025 07.10.2025 30 Mins 07.10.2025 07.10.2025 30 Mins Members Shot demo footage
Creating the Team
video Edit video 08.10.2025 08.10.2025 45 Mins 08.10.2025 08.10.2025 45 Mins Members Edited with captions
Reflect on Team Discussed lessons
project 09.10.2025 09.10.2025 15 Mins 09.10.2025 09.10.2025 15 Mins Members learned
Completing Fill final sections Team Completed
logbook (10–11) 09.10.2025 09.10.2025 20 Mins 09.10.2025 09.10.2025 20 Mins Members collaboratively
Review logbook Team
& video 10.10.2025 10.10.2025 15 Mins 10.10.2025 10.10.2025 15 Mins Members Final check
Team
Submission Submit entries 10.10.2025 10.10.2025 10 Mins 10.10.2025 10.10.2025 10 Mins Members Submitted
2.3 Communications plan
Mode of communication Mixture of face-to-face and online
Once a day for an hour; sometimes twice a day for
Frequency of progress meetings an hour
Shared folder on the cloud; all team members
Responsibility for online documents contribute
Google Meet for online meetings; Google Drive
Tools for communication and Google Docs for collaborative work
3. Problem Definition
3.1 Key localized challenges encountered by our school and
community
Taking attendance manually during each class wastes valuable teaching time and often
results in errors or proxy marking of student’s attendance. Teachers need a quick, reliable
system that works automatically and keeps records safely ensuring to save their and the
students precious time.
3.2 Key issues of concern and their significance to us
Accurate attendance is essential for maintaining academic records and ensuring school
safety. Traditional roll calls are time-consuming, paper registers can be misplaced, and
manual entry often leads to errors. Our goal is to develop a fast, contactless, and secure
system that automates attendance, reducing administrative effort and allowing teachers to
focus on teaching.
3.3 Primary issue identified for focused intervention
Our focus is on developing an AI-driven facial recognition system to automate attendance
marking, eliminating manual intervention and minimizing the risk of human error.
3.4 Problem Statement
How can we help teachers and school staff find a way to automate attendance
recording using AI so that they can save time, reduce errors, and maintain accurate
digital records efficiently.
Self Evaluated Rating - 3/3
4. The Users
4.1 Primary Users and Analysis of Problem Impact
1. Students
● Problem: Attendance is manually recorded by teachers.
● Impact: Students may face delays or errors in marking their presence, and they rely
on teachers to ensure their attendance is correctly recorded.
2. Teachers
● Problem: Teachers must manually take attendance and maintain registers.
● Impact: Valuable teaching time is lost, and the process is repetitive, inefficient, and
prone to mistakes.
3. School Staff
● Problem: Attendance data is recorded on paper and managed manually.
● Impact: Difficulty in organizing, retrieving, and analyzing records efficiently, making
overall school management cumbersome.
4.2 Observations on Users and the Effects of the Problem
The manual attendance-taking system poses several challenges for both students and
teachers, primarily affecting time efficiency, accuracy, and record integrity.
● For Students: Manual attendance consumes valuable learning time and can lead to
errors or proxy attendance, compromising the reliability of records. Mistakes such
as being incorrectly marked absent may go unnoticed due to the lack of
transparency and real-time feedback.
● For Teachers: Taking attendance manually is time-consuming, especially in large
classes, and prone to errors. Maintaining records and compiling reports adds an
administrative burden, while the system offers no tools to analyze attendance
patterns or identify recurring absenteeism.
Implementing an AI-enabled attendance system, using technologies such as facial
recognition, voice identification, or AI-integrated RFID, can effectively address these issues
by automating the process, improving accuracy, and freeing up time for teaching and
learning.
4.3 An Interview Transcripts
Interviewer: Can you share your thoughts on the AI-enabled attendance system you've been using?
User: Absolutely. I’d be happy to. First, I want to express my sincere appreciation for this incredible tool. It’s a truly time-saving way to take
attendance.
Interviewer: That's great to hear! What specifically stands out to you about the system?
User: The accuracy and efficiency are remarkable. Mistakes in marking attendance, especially under pressure—which often lead to disputes or
extra administrative work—are completely avoided.
Interviewer: How do students find this useful?
User: Students no longer need to rely on the manual attendance system, which saves a lot of their time. It’s now a much more reliable and
transparent process.
Interviewer: It sounds like you're genuinely impressed with the model. Any final thoughts?
User: Absolutely. I’m genuinely grateful for the dedication and expertise that went into developing this system. It represents a significant
advancement in technology. Overall, AI-enabled attendance systems offer a faster, more reliable, and transparent alternative to traditional
methods, significantly improving classroom efficiency and accountability.
4.4 Empathy Map
4.4.1 Primary User: Teacher
Says Thinks
“Manual attendance wastes class time.” “Automation would help me start lessons
“Sometimes students mark proxy faster.”
attendance.” “The system should be accurate — false
“I need a reliable system that doesn’t fail positives will cause problems.”
when the internet drops.” “Data privacy is important.”
Does Feels
Marks attendance manually on paper or a Frustrated when attendance takes long.
school app. Worried about errors and proxies.
Cross-checks absentees later. Relieved when technology reduces
Reports discrepancies to admin. workload.
4.4.2 Secondary User: Student
Says Thinks
“Sometimes the teacher forgets to mark “If it works well, it’ll save time.”
me present.” “I hope it doesn’t misidentify me.”
“Face recognition should be quick.” “It should respect privacy.”
“I feel anxious anticipating my turn.”
Does Feels
Walks into class and looks at the camera Curious about AI use.
briefly. Annoyed if recognition fails repeatedly.
Shows ID if facial recognition fails. Happy if the system is seamless and fair.
Checks attendance record occasionally.
4.5 Current User Procedures and Associated Challenges
Students:
● Procedure: Rely on teachers to mark their attendance manually.
● Challenges: Dependence on teachers can lead to errors or missed entries, and the
process consumes time that could be spent on learning.
Teachers:
● Procedure: Call out student names, mark attendance in registers, and maintain
records manually.
● Challenges: Time-consuming, repetitive, and prone to mistakes, especially in large
classes. Manual records can be misplaced, and compiling reports or resolving
disputes adds administrative burden.
School Staff:
● Procedure: Organize, store, and retrieve paper-based attendance records.
● Challenges: Difficulty in maintaining accurate and organized records; lack of
analytical tools to monitor attendance patterns or identify consistent absenteeism.
Overall Observation: Manual attendance processes are inefficient, error-prone, and lack
real-time transparency. These difficulties create a need for an automated, reliable system
that reduces administrative effort while improving accuracy.
4.6 Problem Statement
Teachers and school staff are experiencing issues with manually taking and managing
attendance today because of time-consuming roll calls, errors in marking, and the
difficulty of maintaining accurate paper-based records.
Self Evaluated Rating - 3/3
5. Brainstorming
5.1 Ideas
➔ Facial Recognition: Automatically identifies students in the classroom to mark
attendance without manual effort.
➔ Voice Recognition: Uses student voices to verify presence during class.
➔ AI-Integrated RFID Systems: Tracks students via RFID tags for automated
attendance logging.
➔ 2D Barcode
➔ Computer Vision: Detects and monitors students’ presence in real time through
video feeds.
➔ Machine Learning Analytics: Analyzes attendance data to identify patterns,
irregularities, or trends.
➔ Smart Notifications: Alerts teachers and students about absences or discrepancies
automatically.
➔ Predictive Modeling: Forecasts absenteeism trends and helps staff plan
interventions proactively.
5.2 Priority grid
High Impact / High Impact /
High Feasibility Low Feasibility
AI Facial Recognition Attendance System RFID Smart Entry
Low Impact / Low Impact /
High Feasibility Low Feasibility
Digital Entry App Voice Recognition Attendance
5.3 Optimal AI Solution Selection Based on Priority and
Feasibility
After analyzing the priority grid, the AI solution that best fits our users and is feasible for
our team to implement is facial recognition-based (CV) attendance automation.
Reasons:
● High User Impact: Directly addresses the main problem of manual attendance for
students, teachers, and staff.
● Technical Feasibility: Can be implemented with available tools, open-source
libraries, and moderate computing resources.
● Accuracy and Reliability: Provides precise and real-time attendance recording,
reducing errors.
● Time Efficiency: Saves significant class time for teachers and students.
● Scalability: Can be expanded to multiple classrooms or even the entire school with
minimal adjustments.
Self Evaluated Rating - 3/3
6. Design
6.1 User Workflow with the AI-Enabled Attendance System
1. Students:
● Enter the classroom as usual.
● The system automatically scans faces (via camera or integrated device) to mark
attendance.
● Receive confirmation of attendance (optional notification or display).
2. Teachers:
● Ensure the system is running at the start of class.
● Monitor the automated attendance process, if needed.
● Access real-time attendance records and generate reports instantly.
3. School Staff:
● Maintain the AI system setup and database.
● Retrieve, organize, and analyze attendance data digitally.
● Use insights for reporting, tracking patterns, and administrative decisions.
Self Evaluated Rating - 3/3
7. Data
7.1 What types of data are required to train the proposed AI
solution?
The system primarily depends on facial image datasets paired with accurate labels and
contextual metadata. Collecting diverse and well-structured data will ensure the AI model
performs reliably across different classrooms, lighting conditions, and student
appearances, ultimately enabling a fast, contactless, and accurate attendance
automation system.
7.2 Where and how will the data necessary for training the AI
model be sourced?
Data Needed Where will the Who owns the Do you have Ethical
data come data? permission to considerations
from? use the data?
Have
Sample facial
images of
volunteers Captured locally Yes – verbal and Respect privacy, use
(students & using project Project team / written consent only for attendance
teachers) camera setup school volunteers obtained model, store securely
Want/Need
Additional images Controlled image Yes – participants Avoid bias by including
under varied capture sessions in consented for diverse samples; no
lighting / angles lab Project team testing purposes public sharing
Nice to have
Public facial Yes – licensed for Cite sources properly;
datasets (LFW, Open educational Dataset creators / research and exclude non-
FaceScrub) sources online institutions educational use consensual data
Generated using
Synthetic/ augmentation tools
augmented faces (rotation, Project team Yes – generated Ethically safe – no real
for robustness brightness, noise) (created data) in-house identities used
8. Prototype
8.1 Which AI-based tools and methodologies will be adopted
for constructing the prototype?
To build the AI-enabled facial recognition attendance system, the following key tools
and frameworks will be utilized:
● Python: Primary programming language for implementing AI and computer vision
algorithms.
● OpenCV: Used for image processing, face detection, and real-time video analysis.
● MediaPipe will be utilized for real-time face detection, offering lightweight and
efficient performance on devices such as the Raspberry Pi 5. It ensures that only
valid faces are sent for recognition, reducing unnecessary computation.
● CompreFace API will handle facial recognition and verification, comparing
captured faces against a local database of enrolled users. Its REST-based
architecture allows offline deployment and secure communication between
modules.
NumPy & Pandas: For data preprocessing and management of image datasets and
metadata.
● Face Recognition Library: For facial encoding and matching with high accuracy.
8.2 What AI technologies and software platforms will be
employed to design and implement the proposed solution?
To develop the AI-based facial recognition attendance system, a combination of artificial
intelligence and computer vision tools will be used, including:
● Python – Core programming language for AI model development.
● OpenCV – For image capture, processing, and facial detection.
● MediaPipe will be utilized for real-time face detection, offering lightweight and
efficient performance on devices such as the Raspberry Pi 5. It ensures that only
valid faces are sent for recognition, reducing unnecessary computation.
● CompreFace API will handle facial recognition and verification, comparing
captured faces against a local database of enrolled users. Its REST-based
architecture allows offline deployment and secure communication between
modules.
● SQLite / JSON: To securely store attendance records and facial data.
8.3 What types of decisions or results will the AI solution
produce, and what follow-up measures must be undertaken
once these decisions are made?
Decisions or Outputs Generated by the AI System and Subsequent
Actions
● Decisions / Outputs:
○ Attendance Status: Automatically marks students as present, absent, or
late.
○ Alerts: Flags irregularities such as unrecognized faces or repeated
absences.
○ Reports & Analytics: Generates daily, weekly, or monthly attendance
summaries and identifies patterns.
● Subsequent Actions:
○ Teacher Verification: Review flagged cases for accuracy.
○ Administrative Follow-up: Update records, inform parents, or address
absenteeism.
○ System Updates: Retrain the AI model to accommodate new students or
changing conditions.
Self Evaluated Rating - 3/3
9. Testing
9.1 Which user groups participated in the evaluation and
testing of the prototype system?
The prototype of the AI-enabled facial recognition attendance system was tested by the
following user groups within the school:
● Students: To verify that their attendance was accurately recorded automatically and
to assess the system’s usability from the learner’s perspective.
● Teachers: To evaluate the efficiency, accuracy, and reliability of automated
attendance marking, as well as the ease of accessing and managing attendance
records.
● School Administrative Staff: To test data storage, retrieval, report generation, and
overall integration with school management processes.
9.2 What observations were made regarding user
interactions and experiences during the testing of the
solution?
● During the testing of the AI-enabled facial recognition attendance system, the
following observations were made regarding user interactions and experiences:
● Students:
○ The automated system accurately recorded attendance without requiring
manual intervention.
○ Students appreciated the time saved and the transparency of the attendance
process.
○ No significant usability issues were reported, indicating the system was
intuitive and unobtrusive.
● Teachers:
○ Teachers experienced a significant reduction in time spent taking
attendance.
○ Real-time access to attendance records improved efficiency and minimized
errors.
○ Teachers noted the system provided a reliable method for monitoring
student presence.
● School Administrative Staff:
○ Staff found the digital storage and retrieval of attendance data convenient
and organized.
○ The automated generation of reports simplified administrative tasks and
enabled better tracking of attendance patterns.
○ Minimal technical support was required, suggesting the system was stable
and user-friendly.
Self Evaluated Rating - 3/3
10. Team Collaboration
10.1 How did you actively work with others in your team and
with stakeholders?
● Conducted regular online and offline meetings with team members for
brainstorming and progress review.
● Divided tasks among members according to individual strengths and areas of
interest to ensure efficiency.
● Held frequent discussions with the teacher-in-charge to receive feedback and
make necessary improvements.
● Organized interactive sessions with parents and peers to collect suggestions and
incorporate new ideas.
● Each team member and stakeholder participated in testing and evaluating the
prototype to ensure accuracy and functionality.
11. Individual Learning Reflection
Hrichik Mazumder
“The AI Attendance System project provided invaluable experience in applying AI to real-
world problems. I learned about face detection and recognition, integrating tools like
MediaPipe and CompreFace, and the importance of modular programming. The project
also highlighted ethical AI practices, and improved my collaboration skills. Overall, it taught
me to build working AI prototypes and sparked my interest in computer vision, embedded
AI, and automation.”
Aditi Shukla
“Working on the AI-enabled attendance system helped me understand how design
decisions directly impact user experience and system usability. It also taught me a lot about
how design and technology come together. I learned to think from a user’s point of view
and to make designs that are not just good-looking but also practical , easy to use and
convenient ”
Prince Singh
“By designing the AI-enabled attendance system, I learned how to combine technology and
creativity to solve real-world problems. I gained a deeper understanding of how AI
automates routine tasks and improves accuracy. This project taught me the importance of
user-centered design, clear data visualization, and seamless user experience. I also learned
to collaborate effectively with technical teams and adapt design choices based on
feedback. Overall, it strengthened my skills in innovation, critical thinking, and practical
design application”
Ishanika Mukherjee
“One of the things that surprised me, being a part of this project, is how many people are
actually enthusiastic about artificial intelligence, and only seem unsure regarding it when
they don’t have enough information or understanding. Once the users had been briefed and
explained our project, they were all quite enthusiastic about it.”
Astha Agarwal
“Being the outreach coordinator in this AI project taught me how important communication
and teamwork are in bringing a technical idea to life. I learned how AI can make systems
like attendance more efficient, and I also improved my planning and coordination skills. It
was a really meaningful and confidence-building experience.”
Ishwani Banerjee
“Being part of this project taught me how much communication and collaboration matter in
bringing an idea to life. As the communication and outreach coordinator, I learned how to
connect with people, share updates clearly and concisely, and keep everyone on the same
page. My role as an information researcher helped me dive deeper into how AI systems like
the attendance model actually work. Overall, it made me more confident in balancing both
the technical and human aspects of a project.”
Simran Singh
“Through my involvement in the project I gained valuable knowledge and experience about
artificial intelligence and its uses in daily life. This project has given me the opportunity to
enhance my communication skills. This project also strengthened my understanding of
system design, automation and AI ethics. Overall, this capstone project improved my
technical expertise, analytical thinking, and confidence in developing innovative AI-based
solutions.”
