Top News

During an internship at technology company Affina, Duy Anh observed that attendance and working-hour tracking still involved considerable manual work. He began exploring whether data and AI could help automate the process.

From real-world problem to Edge AI solution

Over four weeks, the teenager independently developed a smart attendance system, working through problem analysis, system design, programming, hardware integration, testing, and deployment.

The system uses a camera connected to a Raspberry Pi 5, with OpenCV and MediaPipe to process images directly on the device using edge computing. Recognition results are converted into structured data and synchronized via an API with a web-based administration system for storage, monitoring, and analysis.

Duy Anh tested and optimized the system to balance processing speed and recognition accuracy within the Raspberry Pi 5’s computing limitations. He also explored anti-spoofing measures, gaining experience with challenges involved in deploying AI systems in real-world settings.

After four weeks of development and testing, Affina recognized the system and installed it for use at the company.

Duy Anh (2nd, R) with family at Rabun Gap-Nacoochee School in the US. Photo courtesy of him

From programming to data analytics

The project also strengthened Duy Anh’s interest in data analytics.

During development, he learned how images captured by a camera could be converted into structured data for storage, querying, and analysis. While programming enables systems to collect and process information, data analytics can help turn that information into insights for operations and decision-making.

The experience reinforced his interest in a field that combines mathematics, statistics, programming, and data-driven problem-solving.

By starting with an everyday operational challenge, he gained experience in identifying a problem, collecting and analyzing data, testing possible solutions, and developing a system with a practical application.