• 💡 The Transition Probabilities of all human beings are almost the same, but the reward metric makes us who we are.

• 🎓 I completed my master's in Robotics and Autonomous Systems, Systems Engineering, from Arizona State University, and my bachelor's from National Institute of Technology Warangal.

• 🤖 I am currently working on Lunar Autonomy Challenge where we build an agent to map and navigate lunar surface in a custom CARLA simulator.

• 👯 I’m looking forward to collaborating on innovative robotics projects and building a community that helps each other grow.

• Computer Science, Artificial Intelligence, Machine Learning, Reinforcement Learning, Deep Learning, Bayesian Learning.
• Computer Vision, Perception in Robotics, SLAM, Motion Planning.
• Programming, Objective Oriented Programming, Data Structures and Algorithms.
• Mechanical Engineering, Finite Element Analysis, Computational Fluid Dynamics, Vehicle Dynamics
• Aerospace Engineering, Orbital Mechanics

May 2024 - Present

• HDR SLAM: Developed a High Dynamic Range SLAM system utilizing event cameras, intensity cameras, and IMU sensors, improving localization and mapping capabilities in high dynamic scene areas by 30%.
• Led my university team in Lunar Autonomy Challenge, where we developed an autonomous agent to navigate, map and analyze the lunar surface in a custom CARLA-based simulator.
• My team Lunar Explorers secured 3rd place in the first round of the challenge. We performed exceptionally in mapping productivity, localization and achieved 75.4% of the top team’s geometric map and rock map score.
• Currently assisting in the design and execution of the Space Robotics and AI course, which is being taught by the mentor, contributing to the development and delivery of engaging, high quality content.

Aug 2023 - Present

• I teach Experimental Physics to undergraduate-level students, aiming to enhance their understanding of theoretical concepts by engaging them in hands-on experiments.
• I facilitate interactive sessions wherein students articulate their queries, allowing me to effectively address and resolve their academic concerns.

May 2018 - Jan 2020

• Enhanced suspension system performance by 30% by leading a team as the technical head of the Suspension department, successfully designing and building an optimized suspension for an All-Terrain Vehicle.
• Increased design accuracy by 45% through modeling suspension and wheel assembly components in CATIA and performing Finite Element Analysis (FEA) in ANSYS, resulting in more reliable and efficient parts.
• Improved compatibility and performance of the front wheel assembly by 35% by leading the design and optimization of critical components, including the hub, knuckle, and spindle, ensuring they met dimensional constraints for steering, suspension, and braking with stock Maruti rims.
• Enhanced roll cage design efficiency by conducting fluid flow simulations in ANSYS Fluent, resulting in a 20% improvement in aerodynamic performance under fluid flow conditions.
• Achieved a 40% increase in component strength and rigidity by conducting iterative design processes and FEA on the front knuckle, incorporating adjustments to steering geometry for optimal performance under dynamic loads.
• Optimized rear wheel assembly design by 25% by considering factors such as pitch circle diameter, power-train coupling, and bearing sizes, resulting in a balanced weight-to-strength ratio with a high factor of safety.
• Increased rear hub durability by 50% through successful FEA to assess its capability to withstand sudden torque transfer, axial push, and bump forces, achieving a robust design with minimal deformation under dynamic loads.
• Reduced vehicle weight by 18% (from 165 kg to 135 kg) while maintaining component strength and rigidity, leading to improved fuel efficiency and maneuverability of the vehicle.
• Improved suspension design accuracy by 50% by implementing programming languages like C++ and MATLAB, resulting in more precise calculations and faster iteration cycles.

Jan 2024 - April 2024

• Implemented Cartographer and gmapping Algorithms within ROS framework on the masterX3 robot platform.
• Conducted comprehensive evaluations to determine the superior SLAM solution by comparing mapping accuracy, computational efficiency, and robustness of Cartographer and gmapping Algorithms
• Enhanced algorithm selection accuracy by comparing GMapping and Cartographer, revealing that GMapping performed 30% better in simpler environments, while Cartographer excelled by 40% in more complex scenarios, highlighting the trade-offs between simplicity and complexity in SLAM systems.

Jan 2024 - April 2024

• Integrated Neural A* algorithm for motion planning in a differential drive robot, utilizing CAD modeling for robot construction and simulation in ROS and Gazebo environments.
• Compared the results of Neural A* with Vanilla A*in terms of percentage of shortest path predictions and ratio of node exploration.
• Improved path-finding efficiency by 35% through implanting Neural A* algorithm, resulting in a 40% reduction in node exploration, 30% increase in shortest path predictions, and 25% faster computation time compared to Vanilla A*.

Aug 2023 - Dec 2023

• Developed a sensor-agnostic model for real-time lying posture detection, utilizing an Arduino with an embedded IMU sensor unit, resulting in more reliable and versatile posture monitoring across various user scenarios.
• Enhanced machine learning model performance by 75% through collecting and processing diverse sensor data from 3-axis accelerometer, gyroscope, and magnetometer for different postures.
• Increased user engagement and real-time feedback by 80% by developing an interactive smartphone interface that communicates seamlessly with the microcontroller, enabling instant posture predictions and providing users with insights on their lying posture.

Aug 2023 - Dec 2023

• Improved real-time object recognition and tracking by 40% by developing an innovative object-counting system that merged Raspberry Pi-4 and Camera Module V2 capabilities with the Faster R-CNN RestNet-50 model.
• Increased versatility of object detection applications by 60% enabling solutions for public safety and traffic monitoring across various scenarios
• Validated system accuracy and reliability through extensive testing in diverse environments, achieving a 95% detection rate across different scenarios and confirming the solution’s solid performance and real-world applications.

Jan 2023 - April 2023

• Enabled a cobot with a camera to autonomously compete against humans in Scrabble, enhancing interactive gameplay and automation.
• Camera Intrinsic and extrinsic parameters were determined to establish the transformations between image pixel coordinates, camera coordinates and world coordinates to enable precise location determination of objects on game board, facilitating the cobot to pick and place the desired objects based on real-time video input from camera
• Developed a robust machine leaning model utilizing Computer Vision, TensorFlow, Keras, and Convolutional Neural Networks (CNN) to recognize English alphabet characters with 90% accuracy and decipher the game board.

Aug 2023 - December 2023

• This project pioneers a real time solution for bridging the communication gap between deaf and hearing communities through sign language detection and translation.
• Leveraging machine learning and computer vision, this system interprets live video feeds into written text.
• It utilizes camera feed, image processing algorithms and a Long Short-Term memory (LSTM) neural network to accurately capture and analyze dynamic hand gestures and movements.
• Implemented Long Short-Term Memory (LSTM) Neural Networks, resulting in a remarkable 50% improvement in model efficiency compared to traditional Neural Networks.

Aug 2023 - Dec 2023

• Enhanced the performance of Pacman agent for diverse grid environments using Q learning, leading to a 30% improvement in agent decision-making in different game scenarios.
• Improved learning generalization by 40% through implementing an Approximate Q-learning agent using feature-extraction, enabling more efficient decision-making across states with shared features.
• Optimized algorithm efficiency by focusing on performance factors such as computation time and scalability, resulting in a 25% reduction in processing time and improved applicability for real-world scenarios.