Topic: Detection and estimation of pothole volume using stereovisionSupervisor: Prof. Indrajit Ghosh, Associate Professor, Civil Engineering• Utilized an IMX219-83 Binocular Camera module in conjunction with NVIDIA Jetson Nano• Calibrated the stereo camera using the checkerboard method and MATLAB Calibrator App• Applied the Block Matching algorithm to create a disparity map and estimated the depth • Detected and segmented the potholes using a pre-trained YOLOv8 (by Ultralytics) modelSkills: Computer Vision · MATLAB · Hardware Integration · Python

• Headed the manufacturing of a Formula Student combustion vehicle within a span of 4 months • Formula Bharat 2023: 8th in the overall event and 3rd in the Cost and Manufacturing event

• Jointly designed a space frame chassis and performed torsional stiffness analysis• Researched carbon fiber in-house manufacturing focusing on monocoque chassis design

• Basics of automotive engineering and manufacturing• Software Skills: SolidWorks (CAD) and Ansys (FEA)

• Researched on fixed wing small scale aircrafts

Topic: A Rear-end Collision Prediction Model Using Machine LearningSupervisor: Prof. Madhumita Paul, Assistant Professor, Civil Engineering• Extracted trajectory data using Kinovea 0.9.5 from two unsignalized intersections• Used surrogate safety measures to assess traffic safety at these intersections • Implemented the K-Means clustering algorithm to classify rear-end collisions • Evaluated various collision prediction models using 7 different ML algorithmsSkills: Machine Learning · Data Extraction & Analysis · Python