Rahul | ME20B145 | Indian Institute of Technology Madras | PR/22/ME/25/145

EDUCATION & SCHOLASTIC ACHIEVEMENTS

Program Institution CGPA/% Year of Completion
Mechanical Engineering|IDDD1 in Robotics|Minor AI & ML Indian Institute of Technology Madras, Chennai 8.1 2025
Class XII (CBSE) Sumermal Jain Public School, Janakpuri 91.4% 2020
Class X (CBSE) Kendriya Vidyalaya CRPF Bantalab, Jammu 84.6% 2018
RELEVANT SKILLS & COURSEWORK
Manufacturing Processes Automation in manufacturing Materials and Design Kinematics and Dynamics of Machinery
Design of Machine Elements Finite Element Analysis Control Systems Introduction to Field and Service Robotics*
Programming Languages Python, C/C++, LaTeX, Java, HTML, JavaScript, SQL
Python Libraries PyTorch, TensorFlow, PySide6, RLlib, Numpy, Pandas, OpenCV, Matplotlib, Sklearn, Seaborn
Software and Frameworks ROS2, ANSYS, Gazebo, Matlab/Simulink, Fusion360, AutoCAD, GeoGerba
PROFESSIONAL EXPERIENCE
4D Gaussian Splatting Guide: Prof. Kaushik Mitra, Computational Imaging lab at Indian Institute of Technology (IIT) Madras.
for Urban Scenes ● Reviewed Urban Street rendering and SLAM3 with Gaussian Splatting for insights on rendering 3D dynamic scenes
Dual Degree Project ● Reconstructed 3D dynamic road scene using SfM4, and LiDAR along with POSE estimation of dynamic objects.
(Aug 24 - Present) ● Developed a segmentation model to separate dynamic objects from static backgrounds and estimate their depth.
EH-Group Developed Electric Vehicle HIL5 to analyze control action of Hydrogen Fuel Cell Controller and process monitoring app.
Systems and Control ● Simulated EV mechanical and electrical dynamics using Simulink model of throttle, brake, ECU and motor module
Intern ● Flashed FCV6 model on Raspberry Pi, enabling cloud/local data exchange of electrical and dynamics parameters.
(Jun 23 - Aug 23) ● Developed PySide6-based HMI7 for system monitoring, to switch between InfluxDB, Kvaser CAN, and RedisDB.
Analyzed casting rejection, identified operational tolerances, and improved system monitoring visualization.
● Developed a KNN model to predict casting rejections using past data, along with process monitoring visualization.
● Used ARMAX and Naive Bayes model to predict convergence toward low or high rejections with 75% accuracy.
MPM Infosoft ● Estimated 2-σ confidence bound over sand properties that lead to low rejection in Principle component space.
Data Analyst
Automated AFN calculation by estimating particle size distribution in images using image processing algorithms.
(May 24 - July 24)
● Pre-processed image with grayscale conversion, morphological operation, adjusting contrast and shadow removal
● Developed the Watershed and distance transform algorithm for particle segmentation and feature extraction.
● Validated results against traditional sieve analysis, achieving error less than 6% of experimentally estimated value.
PROJECTS
Measurement, Designed Segway with PID control, integrated sensors with a Kalman filter, and tuned stability in Simscape Multibody.
Instrumentation and ● Designed a two-wheel differential-drive mini Segway robot using PID controller with a settling time of 4 sec.
Control ● Implemented Kalman filter integrating gyroscope and accelerometer output for better tilt angle estimation.
ME2400 (Jan 22 - May 22) ● Simulated entire model in Simscape Multibody and tuned Kp, Ki, and Kd values based on the results for stability.
Designed & integrated robotic arm in plant layout, to test for ergonomics, safety and production efficiency.
Automation in
● Engineered 3D model and CAD drawing of robotic arm using SolidWorks/Fusion 360, integrated into Tecnomatix
Manufacturing ● Simulated working environment with JACK software, evaluating ergonomics and safety for machinery interaction.
ME3302 (Jan 24 - May 24)
● Compare cost and quality metric of three manufacturing methods to identify the most efficient approach.
Filtered data, modeled dynamic systems, and applied statistical methods to estimate and evaluate key parameters.
Signal Processing of
● Filtered data by removing high-frequency signals through the application of FFT9 from scratch on time series data.
Mechanical Systems ● Developed and simulated data within a multi-degree spring-mass-damper system to analyze its characteristics.
ME7225 (Jan 23 - May 23)
● Applied statistical methods like CRLB, BLUE, and MVUE to estimate parameters and distributions of noisy signal.
Estimated stress factors, validated results, and analyzed stress and strain using Fusion 360 and Python.
Finite Element
● Estimated stress concentration factors for notched plates, with only 9% error compared to experimental value.
Analysis ● Generated triangular mesh of the IITM map using GMSH and estimated the area using 7-point Gauss quadrature.
ME5204 (Jul 23 - Nov 23)
● Analyzed the stress and strain on connecting rod using Fusion 360, refined mesh with mesh.io library in Python.
Trained agents to learn optimal actions in dynamic environments, testing different methods for faster convergence.
Reinforcement
● Applied hierarchical RL with SMDP8 and Intra-option for faster convergence in Gymnasium Taxi environment.
Learning ● Implemented Dueling-DQN9 and MC10 REINFORCE algorithms with and without baseline in Acrobot and CartPole.
CS6700 (Jan 24 - May 24)
● Compared the Q-learning and SARSA for convergence rate over an action stochastic environment with obstacles.
POSITIONS OF RESPONSIBILITY
A competition team of multi-disciplinary students committed to build intelligent, robust, safe autonomous vehicles.
Abhiyaan
● Implemented CAN protocol across vehicle modules and Flash memory-based error handling for fault diagnosis.
Electronics Engineer
● Conducted the workshop “Fundamentals of Power Electronics in EVs”, with participation over 100+ students.
(May 22 - Mar 23)
● Secured 1st position 25+ student teams in IGVC11 competition 2023 conducted by Oakland University, Michigan.
Digital Pen: A pen that digitizes handwritten notes and sketches in real-time conserving the texture of written data.
Electronics Club
● Devised a strategy integrating IMU12, TOF13, CMOS and Force sensors for accurate text and POSE estimation.
Project Member
● Integrated word segmentation with OCR14 to digitize the handwritten text, achieving 67% recognition accuracy.
(May 21 - Mar 22)
● Conducted a workshop on Computer Vision and Robotics with an overwhelming participation of 80+ students.
EXTRA-CURRICULAR ACTIVITIES
● Attained a peak chess rating of 1422, securing a position in the top 5%, while achieving an impressive win streak of 12 games
● Selected for NSO badminton and NCA dance team, completed 8 km Race representing Godavari hostel at IIT Madras in the Dean's Trophy.
|* ongoing|1:Inter Disciplinary Dual Degree|2:Robot Operating System|3:Simultaneous localization and mapping|4:Structure from Motion|5:Hardware in Loop|6:Fuel Cell Vehicle|7:Human Machine
Interface |8:Semi-Markov Decision Process|9:Deep Q-Network|10:Monte Carlo|11:Intelligent Ground Vehicle Competition|12:Inertial Measurement Unit|13:Time of Flight|