IIT Gandhinagar researchers test wearable sensor using machine learning to track sports performance

Researchers at the Indian Institute of Technology Gandhinagar (IITGN) are developing a wearable sensor to track body movements during exercise and sports. The technology is being tested in activities such as weightlifting, strength training, tennis and swimming to help assess how people perform different movements.

The sensor developed by IITGN uses small patches placed on selected parts of the body to record changes caused by movement. These patches will convert physical stretching or bending into electrical signals, which are then sent to a mobile device through Bluetooth for analysis.

Biswajit Saha, associate professor in the Department of Chemical Engineering, IITGN, said, “An experienced coach can often spot such differences quickly. But in a crowded gym or training centre, watching every repetition by every person is difficult. We wondered whether a wearable sensor could provide an additional way to examine movement, even when a coach cannot observe every repetition?”

A machine-learning model is also studying these signals and compares them with movement patterns recorded from experienced athletes, coaches and trainers. The technology aims to help trainers and beginners better understand movement during exercise.

Also read ‘Will disturb peace in classrooms’: Tamil Nadu educationists on questions posed by MHA’s demography

IITGN tests wearable sensor on trainees

Beginners learning a bicep curl may find it difficult to know whether they are performing the exercise correctly without a trainer's guidance. The sensor examines the timing, strength and consistency of movements across repetitions. However, movement differences do not always indicate an error. People may perform the same exercise differently because of their body structure, mobility, or fatigue.

Speaking about developing the sensor platform, Ravi Prakash Verma, a postdoctoral fellow at IITGN’s Department of Chemical Engineering, said, “Our framework is intended to complement conventional coaching by providing quantitative information about movement patterns that may be difficult to assess consistently through observation alone”.

Kosha Navnit Vaishnav, a research scholar in IITGN’s Department of Chemical Engineering, added that the research findings have not yet been published. She added that the team has combined the sensor, data collection and analysis systems into a single platform. The prototype has reached Technology Readiness Level 5 and is under further testing in real training environments. A patent filing is in progress, and the team is refining and validating the system.

The research team has monitored 50 individuals involved in weightlifting and strength training. It has also recorded movements from 30 trainees participating in tennis, swimming and other physical activities to assess the technology across different sports and exercises.

The researchers are also looking to partner with fitness centres, sports academies, trainers and industry organisations to support further testing and product development.