Ongoing Projects

Guides: Prof. Subray R. Hegde & Dr. Sumanth Govindarajan, NITK Surathkal

Developing an AI-powered application that analyses images of failed machine components, specifically shafts, to provide metallurgical failure analysis and actionable recommendations for preventing future failures.
Studied the most common mechanisms of shaft failures in industries and the fracture signatures of each type of failure on the fracture plane. Automated image retrieval from research papers using Linux shell scripting to build training and testing datasets.
Currently conducting a literature review on parameters for data labelling while creating a custom dataset and feature extraction methods for classification of the mechanism of shaft failures.

Projects

GUJCOST Robofest 5.0

Co-designed the mechanical architecture and 3D CAD model of an Autonomous Underwater Vehicle (AUV) featuring modular thruster assemblies, buoyancy control pistons, and a gripper mechanism for object manipulation — submitted for the Ideation Stage of GUJCOST Robofest 5.0 (Govt. of Gujarat initiative).
Performed structural optimization and component integration for enhanced hydrodynamic stability and load distribution, using iterative design analysis to ensure compactness and manufacturability of the AUV chassis.
Drafted comprehensive technical documentation detailing subsystem schematics, bill of materials, and mechanical design workflow, forming the foundation for Proof-of-Concept fabrication and future control system integration.

MT253: Principles of Extractive Metallurgy

Designed and implemented a sustainable process to convert discarded PET bottles into low-cost customizable tiles, serving as a proof-of-concept for decentralised circular waste management systems.
Recycled Polyethene Terephthalate/PET tiles were made using thermoplastic compression moulding. Waste PET bottles were shredded to 16 sq. cm., melted at 300°C, and poured into preheated stainless steel moulds under compressive load.
Fillers were added to enhance mechanical and aesthetic properties, followed by rapid water quenching to obtain durable tiles, thus converting PET waste into a functional product.

IEEE - NITK Chapter

Studied the working principles of Electron Microscopy (TEM & SEM), X-Ray Diffraction (XRD). Explored the theory of energy bands in crystals and solutions of the Schrödinger Equation for specific cases: free electrons, bound electrons, tunnelling, and solid states.
Learnt simulation tools like ASE, VESTA, and abTEM to integrate computational methods for analysing electronic properties of materials and their structural representation, enabling cost-effective research independent of physical equipment and resource constraints.
Conducted simulations on materials such as GaAs, Re₆Se₈Cl₂, cubic BAs, SrTiO₃, Ag, Au, Hg, and conductive polymers (e.g., Polyaniline, PEDOT), computing properties like band structure and density of states.

IEEE - NITK Chapter

Implemented Pure Pursuit Control (PPC), Bicycle Model (BM) & Model Predictive Control (MPC) algorithms for extraterrestrial rover systems, enhancing predictive capabilities and real-time adaptability. Reviewed how self-driving tech can be used to explore distant planets or map uncharted terrains in space.
Compared the performance of the three models under different scenarios, including dynamic obstacle avoidance, response time, computational cost and scalability.
Exploring a deep learning approach using the Transformer architecture for autonomous systems to track moving targets.

IEEE - NITK Chapter

Developed a robust vehicle Number Plate Recognition system [NPR] for Indian vehicles.
Learnt the basics of Machine Learning, including Regression, Classification, Clustering and Neural Networks.
Worked on implementing a YOLO Model to snip the boundaries of the license plate from the given picture of a vehicle and fed it to a custom-built CNN Model for feature extraction, followed by OCR-enabled text extraction.
Deployed TesseractOCR and EasyOCR for efficient number plate text extraction.

Team AeroNITK

As a member of the Structures Team@AeroNITK (2024-2025), I researched different types of Landing Gears, their design and role, to improve their impact on our aircraft dynamics, both mid-flight and during landing.
Contributed towards the fabrication of an autonomous drone, designed and developed as part of Society of Automotive Engineers India - Southern Section (SAEISS) Autonomous Drone Development Challenge (ADDC) 2026.
Won second runners-up (third prize) out of 80 teams for the best payload dropping system - Certificate.