top of page
RESEARCH PROJECTS
DONE & DUSTED (RECENT TO PAST)
1/4
THE RC CAR DRIVES AUTONOMOUSLY ON THE TRAINED MAP
CLEAR DEMONSTRATION VIDEO COMING SOON
Project | 01
Autonomous Downscaled Model Car Using Neural Networks and Machine Learning - (After College- June 2018 to August 2018)
This project was done to prove my interest on autonomous cars and show I have what it takes to do research on the subject
Hardware : Raspberry Pi 3 Software/Language: Python
Technology: Neural Networks
Domain: Robotics - Autonomous Cars
Role Played: Team Leader
Components used: Raspberry Pi 3 Model B, Pi Camera Module, L293D Breakout Board(motor driver),Toy RC car, Lithium ion Battery pack
This Project was presented at IEEE SS12 2018 prelims event held at Jeppiar Institute of Technology, Chennai.
The work includes design of an RC toy car, that can drive autonomously on a given map in which it has been trained rigorously. It involves training the car on the map while capturing photos of the environment and co-relating them to generate a neural network model using machine learning. The RC car has driven autonomously on the given map successfully. Obstacle avoidance has been implemented using ultrasonic sensor and the RC car successfully stops when it faces an obstacle. Haar-cascade image processing has been used successfully to identify stop signs and red traffic signal, so as to make the car stop.
1/3
TESTING & PRESENTATION at
Anamalai Tiger Reserve, Pollachi, Tamil Nadu, India
CAMCALT (Complex Animal Movement Capture And Live Transmission): Project Presentation Complete and was successful.
Dignitaries present were Chief Conservator of Forest Mr.Ganeshan S ,Forest Ranger Mr.Krishnaswamy and Forester Mr.Muniyandi P.
This wouldn't have been possible without the support and encouragement of Mr. Era Meyappan, President,ExNora Naturalists' Club,All India and Mr. T S Jayakumar, President,ExNora Naturalists' Club, Coimbatore. Special thanks to Mr.A Ramesh Secretary,ExNora Naturalists' Club, Coimbatore without whom this project would never have been initiated.
1/5
Project | 02
CAMCALT (Complex Animal Movement Capture and Live Transmission) - Funded Project - (Aug 2017 to Mar 2018)
This project was funded by ExNoRa Naturalists Club for 300 USD and the prototype was presented to the forest officials of the Anamalai Tiger Reserve, Tamil Nadu, India.
Advisor: Dr. Azhagumurugan Ramamurthy, Sri Sairam Engineering College
Hardware : Raspberry Pi 3 Software/Language: Python
Technology: IOT - Open Computer Vision- Image Processing
Domain: Embedded Systems/Automation-IOT
Role Played: Team Leader
Components used: Raspberry Pi 3 Model B, Pi Camera Module, Neo 6M-0-00-1 U-Blox GPS Module,passive infrared sensor (PIR sensor) -Motion sensor, Lithium ion Battery pack
Complex Animal Movement Capture And Live Transmission (CAMCALT) is a device specifically tailored to the need of a forest surveillance and monitoring system. CAMCALT aims to eradicate the existing disadvantages of the Forest Trap Camera’s and integrate the latest technologically developed “Internet of Things” (IoT) into the device for Live feed wirelessly from any part of the world. CAMCALT is used to monitor the forests, animal movements and Poacher Detection. The device uses BCM2837 Integrated chip as the base computer that does the job for the device. With a motion sensor, camera module, GPS and on-board Image Processing Algorithms attached, CAMCALT is an intelligent over watch for the forest landscape. In large numbers, you can have the whole forest under your control.
The most special and powerful feature of CAMCALT is Poacher Detection. This Feature uses the Image Processing techniques to detect the Poacher using the 4 cores of the powerful BCM2837 Processor powered by 1 GB RAM. By using HSV to split-analyze the images and Haar Cascade Detection to match the existing samples with the testing image.
Project | 03
Project | 03
Acoustic Wireless Power transfer using Infrasonic Sound - (Dec 2017 to June 2018)
Advisor: Dr. Azhagumurugan Ramamurthy, Sri Sairam Engineering College
Hardware : Transducers
Software/Language: MatLab
Technology: Wireless Power Transfer
Domain: Electrical Power Transmission
Role Played: Team Member
Components used: Laptop(for simulation),Piezoelectric Transducer, Frequency/Impedance Circuit, Frequency Converter
The project is aimed at wirelessly transferring electric power by means of infrasonic sound using power convertors, transducers and rectifiers and powering a load at a minimum distance of 10 metres. A simulation of the same has been completed.
-
Wireless power transfer is very important to empower our life style Its simplicity and efficiency will revolutionize the wireless power transfer in acoustic field.
-
The technology currently used for acoustic power transmission uses ultrasonic sound of transmitting power. However, the usage of the ultrasonic sound has its own limitations such as low distance of transmission and higher attenuations.
-
We are using infrasonic sound for transmitting power, as infrasonic sound possess better transmitting efficiency over longer distance.
DEMONSTRATION VIDEO OF SMART BRANCHED SWITCH
Project | 04
Design and Development of Smart Branched Switch for Home Automation Systems - Final Year Dissertation (Sep 2016 to Aug 2017)
Advisor: Dr. Azhagumurugan Ramamurthy, Sri Sairam Engineering College
Hardware : Arduino Uno
Software/Language: Embedded C
Technology: IOT - Home Automation
Domain: Embedded Systems/Automation-IOT
Role Played: Team Leader
Components used: Arduino Uno Rev 3, HT12E encoder - HT12D Decoder, RF Transmitter - Reciever module, Bluetooth HC-06 Module, Lithium ion Battery pack
This Project was presented at IEEE SS12 2016 prelims event held at Kongu Engineering College, Erode.
The work includes design of a prototype of smart control system for all electrical and electronic appliances in a house that will require minimal installation while offering a more comprehensive power monitoring and control.
The control of the home appliances is achieved by the use of a single microcontroller, it has a branched structure with the microcontroller as the root memory and many RF transmitters as branches which will control many home appliance .The user interface is via an android application that runs on a smartphone that connects to microcontroller via bluetooth.
The application ensures that the power is managed efficiently by indicating to the user, the connections that are turned on unnecessarily. This helps out in reducing the cost of and also saves energy.
IN PROGRESS (RECENT TO PAST)
1/8
Project | 05
Augmented Reality Based Smart Glasses for Aid of People with Impaired Hearing (July 2018 to Present)
This is a humanitarian project aimed at helping deaf people
Hardware : Arduino Pro Mini
Software/Language: Embedded C
Technology: Augmented Reality
Domain: Embedded Systems-IOT
Role Played: Team Leader
Components used: Arduino Pro Mini, RSI 0.96 inch 4 Pin OLed display module, HC-06 Bluetooth module, small LiPo battery, 4 MAX9814 MIC amplifier modules
The objective of the project is to achieve augmented reality based smart glasses to assist the deaf people on their day-to-day life so as to eliminate the disadvantages that comes with use of sign language which might be tedious for other people who are trying to communicate with them.
Also sounds that we hear on a day-to –day basis such as car honks, ambulance ,etc are identified so as to make the person aware of their environment. This is achieved by creating a library of sounds that we encounter in daily life as well as by using Google’s Cloud Speech-to-Text Application Programming Interface (API) .
The user-friendly UI is via an application that runs on an Android smartphone.
Project | 06
Application Of Uav For Pest, Weeds And Disease Detection Using Open Computer Vision (June 2018 to Present)
This project is being done at Hyoristic Innovations, a drone research startup at Chennai. I am part of the team doing this project and research on it.
Hardware : Nox R2.19 FLight Controller
Software/Language: Python
Technology: Computer Vision - Image Processing
Domain: Robotics - Quadcopter
Role Played: Team Leader
Components used: Nox R2.19 FLight Controller, Logitech HD PRO Webcam C920 , RF Joystick , LiPo battery
The objective of the project is to locate regions that are affected by diseases and pesticides so that we can particularly focus on the regions that are affected and apply chemicals only in that particular area. For this purpose, we divide the entire area into n x n segments and using image processing, the segmented areas are analysed and processed using open CV.
A farmer can easily monitor the area after initial plantation and irrigation to check which area of their lands is being affected by disease, pests and weeds.
Here, a camera, interfaced with a UAV and is used to keep our area under constant surveillance.
PATENTS FILED (Pending)
1. “Development Of Flexible Autonomous Car System Using Machine Learning and Blockchain” under Intellectual Property India Patent – Application No : 201841045731
2. "CAMCALT(Complex Animal Movement Capture and Live Transmission)" under Intellectual Property India Patent – Application No : 201841045515
3. "Design and Development of Smart Branched Switch for Home Automation Systems" under Intellectual Property India Patent – Application No : 201841045730
4. "Design and Development of Acoustic Wireless Power transfer using Infrasonic Sound" under Intellectual Property India Patent – Application No: 201841045516
The abstract with title and name of inventors of the above patents have been published under "OFFICIAL JOURNAL OF THE PATENT OFFICE" - The Patent Office Journal No. 49/2018 Dated 07/12/2018 (This can be viewed for verification).
To see more or discuss possible work let's talk >>
bottom of page