Anish Pratheepkumar Email & Phone Number

Steyr, Upper Austria, Austria

Deep Learning and AI based Robotic object grasping and Manipulation.Research on Neural Implicit Representations and Robotic Surface Processing operations.

Vienna, Austria

Robotics, 3D Vision (perception) and AutomationRobotic Surface Processing Operations.Deep Learning and Neural Implicit Representations.

Stuttgart, Baden-Württemberg, Germany

Predicting Human Behaviors for Safe Human-Robot Interaction using Deep Learning. Applying - Deep Learning on spatio-temporal data (Python - 90%, Mujoco XML - 10%)○ Developed production environment related human-robot interaction scenarios (not avail-able in State of the Art) in Mujoco Physics engine simulator.○ Developed an efficient deep RNN-GRU network by adapting the State of the Art with innovative ideas for human motion prediction (framework used - TensorFlow) .○ Achieved good results on standard metrics by comparative evaluation with RED baselinenetwork and also on Mogaze (2020) human motion dataset.○ Represented the predicted occupancy region of human as 3D convex hulls, aimed at future convex optimization of robot motion.

Siegen, North Rhine-Westphalia, Germany

Autonomous Track Navigation of Quadcopter using Deep Neural Network- Simulation and Hardware Integration.Applying - Deep Learning and Computer Vision (Python - 90%, LUA - 10%) •Created the simulation environment in Coppelia-Sim and developed a path follower algorithm using Python and Lua codes.•Collected synthetic image data and implemented an efficient CNN for image recognition and autonomous navigation on a track.• The quadcopter pre-assembled with - DJI F450 Flame wheel frame(X configuration), FC - Pixhawk Flight Controller, compass, and accelerometer was verified and did a successful flight with RC control.• Added Optical flow sensor and Lidar sensor for indoor GPS denied navigation.• Robot Operating System(ROS) was configured with the quadcopter for autonomous control, by integrating it with ROS installed Raspberry Pi, using MAVLink Protocol.• Analyzed the generalization property of the trained network by testing it in two never before seen, complex tracks that gave a good performance.Tools & Softwares - Keras(TensorFlow backend), Jupyter Notebook, Numpy, Scikit-learn, Mission Planner, Spyder(python programming), Coppelia-Sim(VREP), ROS.

Dubai, United Arab Emirates

Managed projects as a project engineer with key responsibilities like team management, task planning, scheduling, risk management, cost control, quality control, and client coordination.

Dubai, United Arab Emirates

Thiruvananthapuram Area, India

Managed and Supervised the UT inspection of critical launch vehicle components, prepared inspection reports, and attended NDT review committee meetings.•Advanced Equipment Handled – PCUS, USLT 2000, AUS(Acousto Ultrasonic System)

Doctor Of Philosophy - Phd, Robotics, Automation, 3D Vision

Master'S Degree, Mechatronics

Specialized Subjects - Deep Learning, Robotics, Control Theory, Digital Control, Machine Dynamics and System Dynamics, Actorics.

Bachelor'S Degree, Mechanical Engineering, 87%

Activities and Societies: - Executive Member: Mechanical Engineering Association - Organizer: Auto Expo XPLORA 2014 - UUC (University.