Ganapathi Teja Chirravuri

Online Monitoring & Development of AI-Methods for Defect Classification within Fiber Placement

Project: Detection of Surface condensation on PCB using Neural Networks Responsibilities: • Creating custom Datasets of water-droplets on PCB • Labelling, data augmentation and training the NN model • Evaluating the metrics for predicting surface condensation on PCB Using: Anaconda, Python 3.8, PyTorch, ImageJ, Labelme, U-Net, Semantic segmentation models

Project topic: Optimization of image pocessing algorithms. It is a part of EU funded project IKON, collaborated by Germany and Czech Republic. Technische Universität Chemnitz from Germany and Univerzita Jana Evangelisty Purkyně v Ústí nad Labem (UJEP) from Czech Republic. Using Computer vision and state of the art technologies in PyCharm using python 3.6, the Execution time and Frames Per Second (FPS) of the CV algorithms were optimized for Traffic Light and Traffic Sign recognition; Autobahn Detection.

Person recognition and tracking with Mask R-CNN Technologies used: Deep learning, Computer Vision, PyCharm IDE, Python 3.6

Agile Software Development. Worked in Automation of Regression Tests for a verification module which uses Virtual Simulation. Project: Software development for Autonomous Driving (AD) Responsibilities: 1. Automation of Regression Tests for AD-Verification Module. 2. Define, run & update simulation scenarios for Autonomous driving. 3. Presenting test results to the project stakeholders. Tools: Python, ROS, Jenkins, Scenario Description Language, Simulation s/w.

Internship on, "Machine learning for Automotive MEMS sensors". Used Deep learning for classifying the Bosch Automotive sensors (good/faulty). Tools: Python, Tensorflow/Keras.

Project work: In Artificial Intelligence Multi-layered multi-pattern CPG for adaptive locomotion of humanoid robots • Neuronal networks, Deep learning, Reinforcement learning • Implemented in Python 2.7 using Spyder IDE & V-REP simulator • Adaptive locomotion from optimized Nao Robot Crawling to walking

Project - Embedded systems: In Micro Electro-Mechanical Sensors (MEMS) Efficient Maximum power point tracking (MPPT) for Vibration-Energy harvest Tools: Piezo-electric sensor, B-B Converter, C++ in Visual Studio

Advanced Training in Industrial Automation (PLC- Siemens, ABB, SCADA) Project: Automation of Hydro-electric power plant by Distributed Control Systems

Electrical-Power generation through Speed-breakers • Circuit/Model design and testing using MATLAB-Simulink • Analysis of parametric variations and performance evaluation.

Main subjects: • Image Understanding, Machine learning (Neuro Computing) • Media Coding, Software Service Engineering, AUTOSAR, CAN bus • Seminar- Image segmentation with Mask R-CNN • Formal Specification & Verification

Main subjects: • Computer Vision, Embedded Systems design & Architecture • Digital Signal Processing, OpenCV • Seminar - Artificial Intelligence in Autonomous Driving • Smart sensor systems, Software platforms for Automotive systems.

Post Graduate Diploma in Embedded systems AUTOSAR, CAN, Flexray, Embedded C

Main subjects: Electrical machines, Control Systems, Micro controllers and processors, Power systems, Electrical Circuits, Mathematics. Course leader for Seminar: Analogy of Human body and Electrical networks • Comparing different electrical networks with human body systems • Force-Voltage & Force-Current analogies using control systems