UT Embedded Systems student wins international TinyML Hackathon

The winning project focused on Visible Light Positioning, an innovative technology that determines indoor locations using light sources. The team designed and built a complete positioning system running on a Raspberry Pi Pico microcontroller. Their solution demonstrated how smart AI applications can run efficiently on small, energy-efficient embedded devices.

The jury assessed the solutions based on positioning accuracy, model size, processing time and reliability under realistic conditions. Code Detectives achieved the highest overall score.

The TinyML Hackathon was part of the ACM Europe & 4TU.NIRICT Summer School ‘Systems Meet AI’. During the summer school, students and researchers came together to explore how embedded systems and artificial intelligence can reinforce each other. Through lectures, hands-on workshops and collaborative challenges, participants developed new applications and gained practical experience.

For Rishi, the competition provided an opportunity to apply the knowledge and skills gained during the Embedded Systems master’s programme in a realistic technical setting.

Rishi Bala Sivasubramanian

The challenge was not only about accuracy, but about making AI small, fast and reliable enough to run on a tiny microcontroller.

Rishi says that the success of the project came from finding the right balance between efficiency and performance. The team kept their models lean while maintaining high accuracy and tested the solution extensively on the actual device rather than only in simulation. This approach helped them achieve the highest score across the different evaluation criteria.

“Adarsh brought the machine learning expertise, and I focused on the embedded systems side, so the solution really came together at the intersection of both fields,” Rishi says. “It was the perfect opportunity to apply what I have learned in the Embedded Systems programme. Doing it alongside a good friend made it even more rewarding.”

Rishi’s achievement shows how the Embedded Systems master’s programme prepares students for complex technological challenges. The programme combines embedded systems, artificial intelligence and practical problem-solving skills.

As TinyML enables more intelligent applications on energy-efficient devices, the combination of embedded systems and AI is becoming increasingly important in research and innovation. These technologies play a growing role in areas such as smart sensor networks, industrial applications and energy-efficient Internet of Things systems.

Rishi’s achievement highlights how the Master’s in Embedded Systems prepares students to tackle complex challenges at the intersection of hardware, software and artificial intelligence.