Dhinesh Ramasamy



Chennai, India Current: Co-founder and R&D Lead at Netrin Sports Tech Education: IITMadras

About Me

A research engineer involved in creating R & D solutions powered by Artificial Intelligence to enhance lifestyle and quality of living.

Experience

From Aug 2019 To Present

(0 Years 8 Months)

Healthcare Technology Innovation Centre IIT Madras

Project Engineer

End-to-end system design, development, testing and validation of Minimum Viable Products (MVP) for sports and fitness
IP creation through development of proprietary algorithms
▪ Agile execution of multiple projects leading a 3 member team
▪ Created 3 scalable and functional product designs

From Jul 2016 To Jul 2019

(3 Years 0 Months)

Healthcare Technology Innovation Centre IIT Madras

Project Associate

▪ 3+ years of extensive research on inertial motion sensing for sports training and performance enhancement
Proposed a novel method to obtain ride assessment metrics using a single inertial sensor mounted on the bicycle (patent pending)
Devised an inertial sensor hardware for high impact applications with high resolution data acquisition and processing
Proposed an accurate method for human activity recognition for real-time detection of fall using wrist worn wearable device
Designed a platform to improve tennis serves during training and functional validation of the same on 11 players over 2 months

From Aug 2015 To Jul 2016

(0 Years 11 Months)

Healthcare Technology Innovation Centre IIT Madras

Project Intern

End-to-end system design of a wearable ECG patch for continuous cardiac monitoring
Developed power efficient algorithms for on-board computation of heart rate and respiratory rate for clinical applications

From Mar 2019 To Present

(0 Years 11 Months)

Infillion Chennai

Co-founder & Hardware Design Lead

Headed the hardware design of a finger worn wearable to recognize gestures real-time based on inertial sensor data
Created firmware libraries for ver1.0 of the functional prototype

Education

Tools & Platforms

Achievements & Awards

Analog Devices India, Bangalore ( Jul 2017 )

▪ A national level university design fellowship program that encompasses design of practical and marketable IoT solutions
▪ Proposed and developed a revolutionary virtual physiotherapy platform to address practical limitations with existing solutions and thereby presented a functional business model for the same

National Institute of Technology, Trichy ( Mar 2014 )

▪ An national level robotics design contest requiring multiple line follower robots communicating wirelessly to achieve goals
▪ Designed on-board algorithms for a robot to transmit path information to another blind robot to reach a target

Indian Institute of Technology, Madras ( Jan 2014 )

▪ An electronics design contest for prototyping affordable and practical assistive solutions for the differently abled
▪ Designed a low cost walking stick using ultrasound sensors and haptic feedback for intelligent guidance even at steps and pits

Intel India, Bangalore ( Jul 2014 )

▪ Proposed a gesture controlled IoT solution for home automation and prototyped the same on Intel Galileo platform

Academy of Mathematics, Cuddalore ( Mar 2011 )

▪ Secured 200/200 in higher secondary board examinations

collapse1Three

Projects

IoT based Virtual physiotherapy platform with biofeedback

Analog Devices ( Aug 2016 - Mar 2017 )

▪ Solved the problem with accessibility and affordability in conventional physiotherapy training
▪ Created a game based inertial motion analysis platform with biofeedback using Electromyography (EMG) & Heart Rate (HR) and a doctor’s dashboard for remote observation
▪ Built motion capture hardware, developed body mapping algorithms & game user interface (UI)
▪ Designed overall system architecture and sensor communication gateway with Bluetooth LE
▪ Presented a functional business model for rural and urban setup
▪ Press coverage can be found here and here.

Wearable ECG patch for continuous cardiac monitoring and HRV analysis

HTIC Phase I ( Oct - Nov 2015 )

▪ Solved the problem with continuous cardiac monitoring and detection of abnormalities
▪ End-to-end (hardware, firmware and algorithm) design of an Electrocardiogram (ECG) patch
▪ Robust Heart Rate computation algorithm developed using MATLAB & ported to embedded C

HTIC Phase II ( Jan – Feb 2020 )

▪ Proposed superior HR computation logic outperforming traditional signal processing methods
▪ Implemented the logic in a white label product, Movesense ECG, replacing its built in hardware based peak-detection for better accuracy (300%) that is mandatory for HRV & stress analysis

Real-time human activity recognition and goniometry using inertial sensors

HTIC ( Aug - Oct 2018 )

▪ Developed an integrated software tool for real-time 3D visualization and analysis of locomotion
▪ Proprietary activity recognition algorithm detects sit, stand, run, walk, squat and exercises
▪ 2D distance measurement algorithm using 5 body worn inertial sensors (up to 98% accuracy)

Tennis serve learning platform using inertial sensors and voice feedback

IITMadras (patent pending) ( Nov 2017 - Jan 2019 )

▪ Created a technology solution for trainers and players to master tennis serves on the field
▪ End-to-end system design with wireless sensor devices, real-time analysis & feedback engine
▪ Designed racket mounted sensor and proprietary algorithms for stroke classification, ball impact location detection, racket speed & ball spin estimation
▪ User experience & effectiveness has been validated through a study on 11 players over 4 months

Professional cycling ride assessment using a single inertial sensor

IITMadras (patent pending) ( Jan - Mar 2019 )

▪ Proposed a novel approach to assess professional cyclists performance using single IMU
▪ More insights than expensive professional tools to understand overall riding performance
▪ Implemented embedded logic to compute speed, cadence, pedal force, braking & ride quality
▪ The device has huge potential in providing scalable & affordable ride analysis for cyclists

Hybrid algorithm for fall detection using a wrist worn inertial sensor device

HTIC ( Feb - Mar 2019 )

▪ Proposed a cascaded logic with physics, signal processing & machine learning on 6-axis IMU data
▪ Appended a higher true positive rate (TPR) logic with a lower false negative rate (FNR) model
▪ Algorithm was largely simplified by the introduction of physics based features & accuracy of 92% when trained on a dataset with just 447 instances

Assistive IoT switch board panel

( Nov 2017 )

▪ Designed a modular control panel for home automation & formulated a scalable business model
▪ Novel current sensing algorithm to estimate power consumption of appliance(s) using 1 sensor
▪ Intelligent automation algorithms using ambient sensing, Wi-Fi, Bluetooth, infrared, timer

Multi-object tracking in a multiple camera environment

( Mar 2015 )

▪ An embedded computer vision project to seamlessly track an object across multiple cameras
▪ Implemented using multiple web cameras connected to a Xilinx Zynq Zedboard with ARM core
▪ Proposed distributed computing approache and protocols to share processing load and data

Publications