Revan Kumar Dhanasekaran
I love exploring
I am a CAE Data Scientist for Porsche AG through Kube GmbH Ingenieurbüro. During my two years at Wipro Technologies for a Japanese OEM, working on pre-processing and automating NV(Noise & Vibration) and Crash analysis of vehicles, I learned how ML and VR could transform the field of CAE. That's how I ended up at RWTH Aachen University to do an MSc in AI Applied Computer Science. Since then, I have worked on several projects using machine learning, deep learning and reinforcement learning.
Currently, I am co-developing a software tool that automatically detects uncharacteristic events (anomalies) in crash analysis for Porsche AG.
Education
-
RWTH Aachen University
Aachen, Germany
M.Sc CAME
Oct 2018 - Nov 2021
RWTH Aachen offered a unique opportunity to sit for any lecture which paved way to create my own career path. Hence, I focussed on understanding how AI can be applied in computational science.
-
Anna University
Chennai, India
Bachelor of Mechanical Engineering
June 2012 - May 2016
This course covered different facets of mechanical engineering and gave me opportunities to involve myself in many team activities which improved my management skills.
Work Experience
External CAE Data Scientist
Porsche AG, Weissach (Germany)Co-developing a software tool that automatically detects uncharacteristic events (anomalies) in crash analysis and clusters the anomaly simulations in the optimization routine.
Masterarbeit
Porsche AG, Weissach (Germany)Classifying deformation behaviour in crash analysis using spatial-temporal information from few samples.
Research Intern
Porsche AG, Weissach (Germany)Unsupervised low-dimensional representation of crash simulations using geometric deep learning algorithms such as PointNet++ autoencoders.
Predicting anomalous behaviour in crash simulations using anomaly detection algorithms.
Student Research Assistant
Cybernetics Lab IMA & IfU, Aachen (Germany)Parametric optimization of hot rolling of steels using state-of-the-art reinforcement learning algorithms on the hot rolling simulations compatible with Gym AI.
Development of RESTful Server API with FastAPI which logs the hot rolling simulations in arangodb database posted by the client API.
Student Research Assistant
Institute of Technical Thermodynamics (LTT), Aachen (Germany)Development of a Python script for the analysis of the sensitivity of the ignition delay time to the reaction rate constants of individual elementary reactions.
Development of the known reaction mechanisms from the literature linked to a fuel-specific sub-mechanism using Molecular Dynamics simulation and python.
Research on implementation of latest advancements in first-order saddle point search algorithms to find transition states of chemical reactions at the quantum mechanical level using gaussian.
CAE Engineer - Noise and Vibration Analysis
Wipro Technologies, IndiaDevelopment of Automated CAE Review Tool in ANSA using python scripting with a team of three engineers.
Assembly setup of Instrument Panel, Body in White, Suspension group, Closures and Drive Train for Noise and Vibration(NV) Simulations for a Japanese automotive giant.
Part of the team that successfully implemented a crash pilot project with minimal errors.
Projects
-
Incremental Tour Planner
Reinforcement Learning, TSP -
Gait Analysis using ANN
Normal vs Impaired -
Intelligent Finite Element
Stress Prediction, ConvLSTM -
Denoising of Accelerometer Signals
Convolutional AutoEncoder -
Electric Kart - Erudite
EcoKart 2015 -
Image Captioning
CNN-RNN -
Wind Mill DGR Data Analytics
WebScraping, Plotly