Who’s Keeping an Eye on the Exam?
CheatLock uses cameras and sound analysis to flag possible exam rule violations, leaving teachers to review the alerts and decide what happened.
You tell your friend you will cough thrice during your online exam, and that'll mean the answer is B. Or when the teacher looks at someone else's screen, you will sneak a quick glance at ChatGPT.
In an online exam, these tricks can be difficult to spot. Your teacher is only human, after all, and may have nearly 100 students to watch.
But what if a non-human, extra pair of eyes could help? One peak into another window, and bam!
That is the idea behind CheatLock, a platform being developed by four students at Metropolitan University in Sylhet. It uses cameras and sound analysis to flag possible exam rule violations, leaving teachers to review the alerts and decide what happened.
The team wants to bring exam management, monitoring, and results into one place, giving teachers more support and students a fair chance to show what they know.
From a pandemic problem to a project
The idea grew out of the challenges of online exams during the COVID-19 pandemic. As classrooms moved onto screens, teachers had to assess students while struggling to check whether exam rules were being followed.
For the CheatLock team, that challenge became the starting point for nearly eight months of planning, programming, and testing.
Jubayer Rahman Chowdhury coordinates the project and leads AI development. Aitijya Sarker Atibo handles web and app development, Amanur Rahman Aman leads research and innovation, and Sandid Haque Chowdhury works as the business analyst.
Together, they are building a platform designed to help teachers follow what is happening during an exam and review activity that might otherwise go unnoticed.
How CheatLock works
CheatLock uses cameras and sound analysis to identify activity that may need a teacher’s attention.
It checks whether a student’s face is visible, detects additional faces, tracks head position and gaze direction, and looks for potentially prohibited devices. It can also flag unusual sounds or conversations.
The platform turns these observations into alerts and reports for teachers to review.
But an alert is not proof of cheating. A student may look away to think, turn towards a distraction, or hear a conversation outside the room. Each situation needs context.
The developers say a teacher or exam administrator must assess the alerts before reaching a decision. The system’s role is to draw attention to an incident; human judgment determines what it means.
Building trust in the exam room
For the team, CheatLock’s purpose goes beyond detecting possible misconduct. They want students to trust that their effort will be assessed fairly.
“I believe every honest student deserves a fair and trustworthy examination system, and that is exactly what we aim to build through CheatLock,” said Amanur Rahman Aman.
The developers believe results should reflect a student’s knowledge, preparation and ability. They also recognise that fairness depends on clear institutional rules and careful decisions, alongside the technology used to support them.
Taking the idea beyond online exams
The team hopes to extend CheatLock to classroom exams, computer laboratory tests, admission exams and practice tests.
They aim to let teachers organise assessments, monitor activity and prepare results through the same platform.
Making that work across different settings will require further testing. Camera quality, hardware, internet connections and the number of students taking an exam could all affect performance.
Making a mark at competitions
CheatLock has already earned recognition at several technology competitions.
The team won the Idea to Impact Competition at North South University. At CUET SciBlitz 2026, it finished as first runner-up in the Project Showcase category and placed sixth among 108 teams in the hackathon. It also secured fifth place at DUET CSE Carnival 2026.
These competitions have given the students opportunities to present the platform and gain experience as they continue developing it.
An olympiad without invigilators
The team’s next plans include hosting what it describes as Bangladesh’s first invigilator-free olympiad at Metropolitan University in Sylhet, using CheatLock to monitor the competition.
The planned event would take the project beyond competition showcases and into an assessment setting, where its performance can be examined in practice.
That ambition also raises a question central to CheatLock’s development: how can automated monitoring support fair decisions when an alert may have an ordinary explanation?
The founders hope to introduce CheatLock to institutions across Bangladesh before eventually taking it abroad.
What comes next
The team plans to improve CheatLock’s reliability, test it in real exam settings, and work with educational institutions. Feedback from users will guide further changes.
Protecting students’ privacy, securing their data, and reducing incorrect alerts will be essential as the platform develops. An automated alert alone should never be grounds for disciplinary action.
The founders hope to introduce CheatLock to institutions across Bangladesh before eventually taking it abroad. For now, the task is to show that the platform can give teachers useful information and help them make informed decisions.