Experience

The AirLab at Carnegie Mellon University

• United States
• Research
• 1 - 100 Employee

• Research Assistant

  • Jan 2023 - Present

  Pittsburgh, Pennsylvania, United States • Researched and developed terrain traversability cost-map estimation model from Bird’s-eye-view(BEV) RGB local maps , stereo-derived heightmap and robot velocity to predict cost maps that reflect nuanced terrain interaction properties which improved the local navigation by avoiding high cost paths. • Generated Visual Odometry(VO) from stereo images using learning based approach. Generated disparity map from stereo images, registered RGB pointcloud with VO to generate local RGB maps and… Show more • Researched and developed terrain traversability cost-map estimation model from Bird’s-eye-view(BEV) RGB local maps , stereo-derived heightmap and robot velocity to predict cost maps that reflect nuanced terrain interaction properties which improved the local navigation by avoiding high cost paths. • Generated Visual Odometry(VO) from stereo images using learning based approach. Generated disparity map from stereo images, registered RGB pointcloud with VO to generate local RGB maps and height maps • Worked with Prof. Sebastian Scherer on adaptive & uncertainty-aware perception for off-road driving project; collected & analysed visual, geometric, robot dynamics data & tested learning models on Yamaha ATV. https://theairlab.org/offroad/ Show less



Bharat Electronics Limited

• India
• Defense and Space Manufacturing
• 700 & Above Employee

• Robotics Software Engineer

  • Aug 2019 - Jul 2022

  Bengaluru, Karnataka, India Worked in the development of Lidar, Camera, Sonar and IMU based perception modules for indoor and outdoor unmanned ground vehicles (UGV) • Deployed a Deep learning network for road scene segmentation, with a mIOU of 0.71. The objective was to get road surface from street images for path planning, obstacle avoidance and decision making for outdoor robot navigation • Developed Lidar based SLAM module for 4-wheel drive differential robot to navigate outdoor GPS denied environments… Show more Worked in the development of Lidar, Camera, Sonar and IMU based perception modules for indoor and outdoor unmanned ground vehicles (UGV) • Deployed a Deep learning network for road scene segmentation, with a mIOU of 0.71. The objective was to get road surface from street images for path planning, obstacle avoidance and decision making for outdoor robot navigation • Developed Lidar based SLAM module for 4-wheel drive differential robot to navigate outdoor GPS denied environments. Incorporated sensor fusion of IMU, wheel odometry for localization from extended Kalman filter methods • Designed SLAM module for Autonomous Mobile Robot Sanitizer, a two wheel differential robot, to provide mapping to sanitize indoor spaces with UV-C light giving a sanitisation coverage of 100 sq ft in 5 min • Created an Android application with two modes to remotely perform teleoperation, mapping, navigation and monitor sensors and health of indoor autonomous robot • Configured and tuned move base package to link global and local planner to accomplish navigation task on indoor and outdoor robot platforms • Interfaced 3D Lidar, RGBD Camera, 8 ultrasonic and 8 infrared sensors with Jetson edge computing devices to gather perception data for navigation of four wheel differential outdoor robot • Calibrated inertial sensors -Accelerometer, Gyroscope and Magnetometer with Arduino controller to get data for localization for indoor and outdoor robots

• Radio Frequency Design Engineer

  • Sep 2016 - Jul 2019

  Bengaluru, Karnataka, India Design, testing, and integration of baseband and RF subsystems for Software defined Radio and RADAR * Conceptualized and developed waveform generation module for S- band RADAR in a team of two for generating DAC based modulated waveforms with digital filters with a option of tunable bandwidth between 2.5MHz and 5MHz * Built transceiver modules for L band Software defined Radio (SDR) in a team of two with FPGA controller with automatic gain control and tunable filters maintaining flat… Show more Design, testing, and integration of baseband and RF subsystems for Software defined Radio and RADAR * Conceptualized and developed waveform generation module for S- band RADAR in a team of two for generating DAC based modulated waveforms with digital filters with a option of tunable bandwidth between 2.5MHz and 5MHz * Built transceiver modules for L band Software defined Radio (SDR) in a team of two with FPGA controller with automatic gain control and tunable filters maintaining flat gain and reducing spurious frequencies by 4%



Iotacell

• Summer Internship

  • Jun 2015 - Jul 2015

  Bengaluru, Karnataka, India smart home automation project Interfaced Raspberry Pi with multiple sensors. Controlled appliances such as bulbs and fans via internet by developing an Android app, thus automating appliances through the internet. Also deployed the cloud application using IBM Bluemix