Exercise Recommendation

Key Idea

We are exploring how to use AI and sensors (Pervasive Computing) for better, personalized, Yoga - the holistic Indian health system, by nudging a practitioner towards long-term exercise adherence. We do this supporting learning and practice of Yoga with performance and environment monitoring, and data-based (group) recommendation of poses and breathing.

Quick Start

1. Get a quick project overview from our May 2026 poster. [Poster]
2. Our Surya Namaskar (SN) explorer tool, SN-YE, pronounced "sun-yee", is intended for a beginner to tell about SN, variants, breathing, utterances/ mantras, etc. powered by an SN ontology (SN-YO). Please try SN-YE and provide feedback on the form. [SN-YE Tool] [SN-YO Ontology] [Survey Form] [Paper]
3. Read our paper(2025) on overall idea of personalized SN with technology, and an initial prototype of metadata management, sensor data collection, analysis and platform.
4. Return to the broader Trustworthy Group Recommendation page for team formation and fairness-focused group recommendation work.

Conceptual Overview - SN Personalization

Key Contacts: Hari P Gupta, Biplav Srivastava

To the best of our knowledge, this is the first paper that comprehensively examines decision support issues around Yoga personalization, from pose sensing to recommendation of corrections for a complete regimen, and illustrates with a case study of Surya Namaskar (SN) - a set of 12 choreographed poses.

Representative Publications

• [2025] Technology-assisted Personalized Yoga for Better Health -- Challenges and Outlook
  Vivek Kumar, Himanshu Sahu, Hari Prabhat Gupta, Biplav Srivastava
  [Paper]
  

Modeling for SN

Key Contacts: Mansi Dodiya, Bharath C Muppasani, Hari P Gupta, Biplav Srivastava

The SN Yoga ontology (SN-YO) models Surya Namaskar as asanas, numbered pose occurrences, and sequence variants. It also includes a pose correction layer for the SN-variant followed at IIT BHU (our baseline). This layer models body parts, breathing patterns, safety notes, pose errors, and correction instructions. With information in multiple languages, any user application, like the SN-YE tool, can retrieve SN information in English or other supported languages (currently, Hindi and Telugu).

Representative Publications

• [2026] SN-YO: An Ontology for Modeling Surya Namaskar as a Sequence-Based Yoga Practice
  Mansi Dodiya, Kishan Kumar, Bharath Chandra Muppasani, Biplav Srivastava, Raghava Mutharaju, Hari Prabhat Gupta
  [Paper]
  

Sensor Data Collection and Analysis for SN

Key Contacts: Kishan Kumar, Hari P Gupta, Biplav Srivastava

We have collected wearable sensor dataset from participants performing the 12 postures of Surya Namaskar at IIT-BHU. The data was recorded using Arduino Nano 33 BLE Sense devices and ESP32-C6 smartwatches, mainly worn on the right wrist. The dataset includes motion signals from accelerometers, gyroscopes, and magnetometers. We have data over two years of which a subset is released - from 21 participants between 9 July and 9 September 2026, covering 35 collection days, 107 recording sessions, and 1,098 SN rounds.

Representative Publications

• [2026] Hari Prabhat Gupta, Biplav Srivastava , Gaurav Gaikwad, Himanshu Sahu, Mansi Dodiya, Priya Gautam, Sriyog Holkar, Priyanshi Srivastava, Kushagra Baranwal, Kishan Kumar
  Yoga4All: An End-to-End Framework for Collecting Wearable Sensor Datasets to Recognize and Monitor Surya Namaskar Postures
  IEEE Dataport, September 16, 2026, doi:10.21227/prkc-h546
  [Paper]
  

Funding

This work has been possible due to a VAIBHAV Fellowship (2025-2028; 17 proposals from among 216 selected globally) for summer visit and collaboration with IIT-BHU. This is in additon to support for group recommendation work from USC and SCRA.