Experience

1. Honeywell Technology Solutions | BU: Safety and Productivity Systems

Location: Bangalore, India

Position: Embedded Engineer – II

Duration: July 2019 – July 2022

Implemented a fire detection algorithm into the existing 3IR flame detector, ensuring compliance with critical EN54 (European certification), resulting in a reduction of 10 months in the development cycle and the saving of over 10 human resources for data collection.
Developed a Machine Learning (ML) framework using Python scientific libraries that was compatible with the existing dataset for Triple IR sensor-based flame detection. This framework saved 150 man-hours of the data collection cycle.
Trained and deployed a state-of-the-art object detection model (YOLOv4) on NVIDIA Jetson Nano for autonomous maritime search-and-rescue purposes, which successfully secured project funding from the global team.
Collaborated with a 3-member team to analyze IR flame signature data and design an ML algorithm for deployment on legacy hardware. Achieved 98% flame detection accuracy using Python, MATLAB, and IAR embedded framework.
Mentored 3 interns in the areas of Machine Learning (ML) and Computer Vision (CV), providing guidance on technical aspects and fostering professional efficiency.

Patent No. 20240021059, published on 01/18/2024, in the United States Patent and Trademark Office.
Patent No. 20230408476, published on 12/21/2023, in the United States Patent and Trademark Office.
Filed 6 Trade Secrets and 2 U.S. patent applications, resulting in 8 IP awards.
Awarded the Diamond award for securing the 2nd position out of 276 ideas presented at the annual innovation competition, SParkS.
Awarded the Silver award for developing new Fire Path for flame detector.
Received the Silver award for developing a new Fire Path for the flame detector.
Certified as a Six Sigma Green Belt in DFSS Hardware and completed an AI/ML Bootcamp.

Prototyped a novel visible plus thermal camera-based flame detection system for the annual innovation challenge. This solution utilized Computer Vision (CV), deep learning, and TinyML frameworks, contributing to the creation of high-value Intellectual Property.

Location: Chennai, India

Position: Intern

Duration: May 2018 – July 2018

Designed and modeled a Neural Network using Keras to automate the classification of automobile telltales, achieving an accuracy rate of 98%.
Developed an algorithm utilizing XCode (C++) and OpenCV that was three times faster than traditional methods for computing the degree of rotation of the analog fuel indicator in a car.