Software Engineer Developer

Vijoy Sunil Kumar +1-720-***-****

*****.*****@*****.*** linkedin.com/in/vijoys

github.com/vijoy-sunil

Seeking full time position as an Embedded Software Developer Education

Master’s degree in Embedded Systems Engineering Bachelor’s degree in Electrical and Electronics Engineering University of Colorado, Boulder, 2016 – 2018 Amrita School of Engineering, India, 2009 - 2013 GPA 3.72 GPA 7.85 (10.0 scale)

Current work

Udacity’s Self-Driving Car Nanodegree

• Sensor fusion, localization and control algorithms for autonomous vehicles. Drone Net – Multi spectral drone detection and tracking in shared air space

• Integrate hardware and software to develop a low cost acoustic sensor array to estimate azimuth and elevation of specific sound source. Previous experience

Embedded software engineer

Amrita Centre for Nanosciences, India Aug 2013 – Aug 2015

• Firmware lead responsible for Algorithm design and Firmware development for a smart solar charge controller.

• Developed an Intelligent prediction algorithm for efficient battery and solar charge management.

• Introduced a Software development process and an Efficient Test bench setup. 12V/24V residential solar charge controller C, PIC18F47J53

• Convert any 12/24V normal home inverter into solar inverter and efficiently reduce grid power consumption.

• 7 Operating modes, Plug-n-Play operation, Periodic data logging, USB and Bluetooth user interface. Add on board for 12V/24V Off grid solar inverter C, PIC18F26J50

• Developed SPI driver for flash storage and I2C driver for board to host inverter controller communication.

• Developed Message Layer firmware for efficient means of communication between board to host inverter controller.

• Added graphical LCD, USB and Bluetooth interface for smarter user interface. Expertise

Programing Languages: C VHDL Verilog Python CUDA C++ Assembly 8051 Softwares: Xilinx Vivado Design Suite Libero SoC Design suite Quartus Prime Development Suite ModelSim Octave IAR Embedded Workbench MPLAB IDE MATLAB Kinetis Design Studio Keil CodeBlocks KiCad Proteus ARDUINO IDE Eclipse IDE Anaconda. Worked on: Microsemi SmartFusion Altera MAX10 DE-10 lite, DE-1 SoC NVIDIA Jetson TK1,TX1 ARM cortex M0+, cortex A8 nRF24L01 TIMSP432 Siemens C501 Atmel AT89C51RC2 PIC 18F26J50, 18F47J53 Atmega328 ATF16V8C SPLD Raspberry pi3. Other skills: Machine Learning TensorFlow Keras OpenCV FreeRTOS Embedded Linux POSIX RT APIs Code register-level device drivers SPI I2C UART USB Github Kicad schematics Static code analyzer – Doygen, Splint. Projects

FPGA based RGB color detector Verilog, Altera DE1-SoC 2017 Implemented HPS to FPGA communication to transfer the dominant color data obtained from image captured by USB web camera. Also developed VGA controller in FPGA to display the color data on a VGA monitor. Vehicle detection and tracking OpenCV, Python, Keras 2017 Tracking vehicles in camera images using image classifiers such as SVMs and HOG and using filters to fuse position data. Advanced Lane detection OpenCV, Python 2017

Detect lane lines in a variety of conditions, including changing road surfaces, curved roads, and variable lighting, using OpenCV to implement camera calibration and transforms, as well as filters, polynomial fits, and splines. Behavioral Cloning Keras, Python 2017

Drive a car around the track in a simulator and record steering angles and image data and train a Deep Neural Network to make the car drive autonomously around the track. The model is trained, validated and tested using Keras Traffic Sign Classifier TensorFlow, Python 2017

Built and trained a 5-layer Convolution Neural Network on German Traffic Sign dataset to classify Traffic Signs based on LeNet architecture. Achieved a validation set accuracy of 97.9 % and a test set accuracy of 95.8 %. 6U CubeSat Altitude Determination and Control System C, BeagleBone Black 2017 Implemented SPI driver in BeagleBone Black for interfacing with 4 Reaction wheels for precision pointing of the CubeSat. Developed Error handler task and an efficient Data logging task that can lighten the communication load between ground and spacecraft. GPU accelerated Stereo Imaging OpenCV, CUDA, NVIDIA Jetson TX1, DUO3D camera 2017 Implemented stereo disparity imaging algorithm in GPU to compute depth map from the left and right frames captured using the DUO camera. Achieved 6.5x speed up when compared to CPU processing. Real time sign language interpreter OpenCV, C, NVIDIA Jetson TK1, Logitech C200 USB camera 2017 Recognize 20 of 26 alphabet hand gestures in real time using SCHED_FIFO scheduling policy and processing within a time period of 330 ms for each of the 5 threads implemented.

Contact this candidate