Kiran Gude
Robotics Software Engineer
Roboticist with expertise in Autonomous
Navigation and Deep learning along with
extensive knowledge in Hardware/software Co-
Design.
******@*****.***
Munich, Germany
WORK EXPERIENCE
Master Thesis- R&D
BMW Group (Innovative Automation)
03/2020 - Present, Munich,Germany
Implementing 3D Visual SLAM with end to end spatial AI stack for mobile manipulator using Intel Real sense D435i RGB-D cameras.
Visual inertial odometry with Deep learning-based
features(superpoint) for robust localization & pose estimation with frontend and backend layers.
Metric semantic 3D mapping with Deep learning based semantic segmentation model for Indoor scenes.
Custom training of semantic segmentation on indoor based production plant dataset.
Contact: Lukas Burseg - *****.******@***.**
Sofware Developer
Cisco Systems
03/2016 - 06/2018, India
Development of Board bring up and Linux device drivers for cisco ASR9x series edge router hardware platforms. Testing firmware and various network drivers for
hardware reliability and packet forwarding with other chipsets like ASIC and FPGAs etc.
Implemented Python automation framework for
hardware reliability and module testing.
Software Developer
Samsung R&D Institute,India
01/2014 - 03/2016, India
Development of board bring up for MEMS sensor
chipsets in Samsung MSMx series and Exynos-3
Qualcomm based variant model devices.
Developed Linux device drivers for sensor chipsets & designed hardware abstraction layer for all drivers. Developed deep learning-based activity classification and gesture recognition using accelerometers & gyroscope for mobile & wearables.
Design and Implemented various Sensor Fusion & sensor calibration algorithms for inertial sensors, magnetic sensors, and vision sensors.
SKILLS
C/C++ Python Robotics ROS SLAM
Computer vision Localization Motion Planning
Mapping Deep Learning PyTorch MATLAB
TensorFlow CUDA TensorRT Gazebo
Linux Kernel Embedded Systems
PROJECTS
Autonomous Smart Factory to manage autonomous
mobile robots in Warehousing. (2019)
High level: Task Management, roadmap planning, charging management, finding feasible trajectories and perception of the local environment of automated mobile robots.
Design & Implementation of low-level parts which includes EKF sensor fusion & localization, motion planning, path planning with A- star, path smoothening(Bezier splines) and obstacle avoidance . Evaluating the scaling of robots versus area of the warehouse. Computer vision-based motion planning for a mobile robot system in a dynamically changing grid
environment. (2019)
Implemented image processing algorithms to identify, localize and classify the objects in robot configuration space.
The data related to each and every object stored into a geo- localized database, which is represented as model of the environment.
sampling based motion planning RRT* algorithm is implemented and navigated to goal.
Imitation learning assisted Reinforcement learning of robot to avoid humans during its motion in human-
robot workplaces (2019)
RL Agent learns from demonstration data of crowd motion which contains annotated trajectories and smoothly avoid humans Infront of it during its navigation from one point to point. Using Unity based ML-Agents toolkit for simulation environment and also for training, learning of agent.
Implemented SLAM for mobile robot based on laser
scan sensor 2D grid environment. (2018)
Implemented ROS mapping node which builds map based on laser beam reflectance probability data of the occupancy grid. Implemented particle filter based localization algorithm to localize mobile robot inside map with odometry sensor raw data. ACHIEVEMENTS
Published a paper on “Non-immersive method of
accelerometer calibration for mobile devices” in IEEE Published a paper on “user agnostic error
compensation for MEMS magnetic sensor in handheld
devices” in IEEE.
Achievements/Tasks
Achievements/Tasks
Achievements/Tasks
Page 2 of 2
EDUCATION
Masters in Autonomous system
Technical University Berlin
09/2018 - 11/2020,
Probabilistic Robotics Applied Robotics
Deep learning &
Computer vision
Computer vision &
Multimedia Systems
Applications of Robotics
and Autonomous
Systems
Bachelors in Electronics &
Communication
Jawaharlal Nehru Technological University
India
ACHIEVEMENTS
Published a paper on “FPGA Implementation of
Canonical Signed Digit multiplier” in International Journal of Electronic and Communication Research
(IJECR) Volume 3 No: 2 (2012) PP:73-78.
Granted patent for novel method of Heart rate
measurement by fusing accelerometer sensor and
Microphone data in Samsung Research.
CERTIFICATES
Coursera:Neural Networks and Deep Learning-
Deeplearning.ai (2018)
Coursera: Convolutional Neural Networks-
Deeplearning.ai (2018)
Fundamentals of Reinforcement Learning by University of Alberta & Alberta Machine Intelligence Institute on Coursera (2019)
Reinforcement Learning: Sample-based Learning
Methods (2019)
Reinforcement Learning: Prediction and Control with Function Approximation (2019)
Introduction to TensorFlow for Artificial Intelligence, Machine Learning, and Deep Learning
LANGUAGES
English
Full Professional Proficiency
German(A1)
Elementary Proficiency
INTERESTS
Hiking Cricket DIY Badminton
Courses