Quality Assurance Electrical Engineering

Prasanna Kolar

PhD Candidate, Department of Electrical Engineering,

University of Texas, San Antonio, TX

Email:********.****@*****.***

Tel: 734-***-****

a. Professional Preparation

University of Texas at San Antonio Computer & Electrical Engineering Masters in Science, 2015 b. Appointments

2015-Present Doctoral Research Fellow, University of Texas at San Antonio, TX 2014-2015 Graduate Research Fellow, University of Texas at San Antonio, TX 2009-2013 Quality Assurance Lead, IHeartMedia, at San Antonio, TX 1999-2008 Quality Assurance Lead and Manager, Bank of America, Ann Arbor, MI c. Products

Publications

P Kolar and K. Bhaganagar Detection and tracking gas plume implementing the centroid method using Unmanned Aerial Vehicles, 2016 (ACC 2017 submission)

P Kolar and K. Bhaganagar Implementation of a gas plume detection system on autonomous systems Fluids Conference Presentation 2016

M Merino, P Kolar, Y Huang and D Pack, Unmanned Aerial Vehicle control using Brain Computer Interface 2016

MS. Thesis

Unmanned Aerial systems in Dynamic environments – Implement a Linux-ROS based Decentralized Multiple Robot system that successfully executes Dynamic missions. Used Novel combination of Consensus based bundled algorithm with Subsumption Architecture to implement tasks.

Publications (in review with P Kolar as corresponding author)

K. Rao, P. Kolar, S. Reddy, K. Bhaganagar, Control framework for chemical plume prediction, International Journal of Robotic Research ( submitted June 2016)

P. Kolar and K. Bhaganagar, Chemical and Robotic Laboratory framework for UAV detection of chemical plumes, International Journal of Robotic Research (Submitted May 2016) Projects worked

Gas plume sensing and tracking drones: Implemented (simulation and hardware) technology that will enable a drone platform to sense chemical plumes and navigate. System will be used to detect gas plumes, navigate to the source and trace the plume area. Octocopter & Sensors integrated with Arduino & Linux onboard computer were used to build this system. Code developed on ROS with Python, CPP programming

Executing complex tasks in dynamic environments using drones: Implemented an architecture in ROS that executes tasks on a drone based on subsumption architecture, while allocating tasks using Consensus based bundling algorithms. Swarm of UAVs that executed tasks autonomously in a dynamic environment.

Brain controlled Drones: DOD sponsored project for designing a system of drones that can be controlled by human brain signals. Commands are generated and are used on autonomous drones using decentralized task allocation and execution algorithms. This project was implemented using ROS and uses LabStreamingLayer (LSL) to exchange information between the BCI system and the Drones. We have implemented a multiple quadcopter system that works just by human thought.

Autonomous Multicopter control: DOD sponsored project to Design and implement an autonomous quadcopter system that tracks an object using a camera, maintains altitude using sonar, navigates in a combination of set formations or paths, all done using ROS and python. This project involved building a quadcopter from scratch, assembling components from various vendors onto a quadcopter frame, and developing autonomous programs to control the system. This system uses PX4 Pixhawk autonomous system on a DJI F450 frame.

Autonomous Ground Vehicle control: Develop control programs that uses trajectories and executes them on Pioneer Ground Robots (UGV). UGV autonomously followed trajectories of various shapes and sizes and reached target, using Sonar as a sensor.

Pedestrian tracking computer vision: Implementation of a pedestrian tracking system, using computer vision. This system was implemented using Python programming language. For prediction and noise reduction, Kalman filters were implemented. This system successfully tracks multiple people. This system can be adopted and implemented on Robots, (ground and aerial), to track subjects of interest.

Pedestrian detection and tracking using Neural Networks: Implementation of a pedestrian detection and tracking system that uses the concepts of Machine Learning – Convolution Neural Networks(CNN). System successfully detects pedestrians using training data that is provided. This system uses the state of art Tensorflow machine learning tool and provides accuracy of over 94% at the rate of 15 frames per second.

d. Synergistic Activities

1. Editor, Special Issue, Physics of Fluids (2016), Tribute to Legacy of John Lumley e. Service Activities

(i) Activities at local communities

(Department) Presenter at local schools in San Antonio: Present my Robotics activities and kindle the spark of interest in youngsters from local schools (2014 – Present)

(ii) Professional Societies:

IEEE

e. Press Release

CBS News : http://www.cbsnews.com/news/drones-controlled-by-brain-waves-in- development-in-texas/

UAS Vision : http://www.uasvision.com/2014/09/04/utsa-researchers-study- brain-signals-to-operate-uas/

My San Antonio : http://www.mysanantonio.com/news/local/communities/stone- oak/article/Researchers-try-to-catch-a-brain-wave-5728289.php Summary of Activities in the Last Five Years: P Kolar is currently pursuing an exciting and challenging Doctoral research under prof. Mohammed Jamshidi, world’s leading expert in the area of Controls and Autonomous systems. Kolar had rigorous training as a Master’s student under the guidance prof. Daniel Pack, leading expert on Unmanned Aerial Systems. Kolar has successfully implemented a gas sensing drone architecture, working with prof. K Bhaganagar. Before pursuing the area of Controls and Autonomous systems, Kolar established himself in the area of software development and quality assurance, mastering several programming concepts and languages, standards over several business areas and technologies.

Contact this candidate