IIIT-Hyderabad students unveil innovative ML model-switching approach for real-time traffic monitoring on smartphones

Four CSE second-year students demonstrated a dynamic ML model switching approach on smartphones for real-time traffic monitoring.

Hyderabad: The International Institute of Information Technology (IIIT) Hyderabad students have come up with a dynamic machine learning (ML) model-switching technique on smartphones, enabling real-time traffic monitoring that adapts to changing conditions depending on the traffic flow.

The team comprising undergraduate second-year CSE students – Kriti Gupta, Ananya Halgatti, Priyanshi Gupta, and Larissa Lavanya – under PhD student Akhila Matathammal’s mentorship and guidance of Prof. Vaidyanathan, who is part of software Architecture 4 Sustainability group at Software Engineering Research Centre, worked on a dynamic model switching approach titled EdgeML Balancer, for object detection on edge devices such as smartphones.

---

The concept of model switching refers to using ML models on edge devices such as a smartphone, where the model balancer intelligently switches between models based on the incoming inputs.

This approach, students, not only prototyped on the Qualcomm Innovators Development Kit (QIDK) platform to simulate different scenarios but also evaluated the approach on real-time traffic data using the smartphone.

“Our previous work on model switching was for ML models deployed on the Cloud where we tried to optimise accuracy and latency depending on the situation. But in this case, we experimented on an Edge device, which is resource constrained. Here we extended the self-adaptive switching approach to include sustainability in terms of energy consumed by the model,” said Akhila.

Prof. Vaidyanathan said the prototyping phase helped to optimize strategies for Samsung Galaxy M21 smartphone, on which the primary deployment was tested. “We will now be testing it out on the Samsung S24 Ultra and intend to exhibit it at the upcoming R&D Showcase which will be held at IIIT-Hyderabad in March,” he said,

Students came up with this new approach after they walked into the Embedded Systems Workshop – hands-on course for second students that emphasises learning by doing – searching for a novel project to work on.

Qualcomm, as part of a collaborative initiative to develop EdgeAI use cases at the IIIT-Hyderabad, provided the QID kits comprising hardware, software and customer support built on the latest premium SnapDragon system-on-chip to the team. Students explored and experimented with the kit, resulting in an innovative approach.

Students’ research work has been accepted for presentation at a workshop, which will be held on the sidelines of the International Conference on Software Architecture (ICSA) 2025.

Ananya said that they did not have a paper publication in mind when they initially embarked upon the project. “It was more of an exploratory project for us and what worked in our favour is that everyone in our team was also interested in learning new stuff,” she added.