PROJECTS
AT IEEE RAS
Explore the cutting-edge robotics and automation projects built by the members of the IEEE RAS Student Branch Chapter at NIT Silchar. From autonomous rovers to high-impact research publications, our community actively engages in pushing the boundaries of technology.
Robots under IEEE RAS
Wheeled Rover
Team NITS YANTRARNAV, representing NIT Silchar, has qualified for the live demonstration round at the ISRO Robotics Challenge-URSC 2024 (IROC-U2024) and entered the semi-finals!
Poser
Robot that can mimic human emotions and human posture
6-Axis Robotic Manipulator
Has 6 degrees of freedom → 3 for position (X, Y, Z) + 3 for orientation (roll, pitch, yaw). Structure: base, shoulder, elbow, wrist, end-effector. High flexibility – mimics human arm movements.
IInvenTix Projects
IInvenTix project: A global Showcase project (NEERVAHAN)
Design and Implementation of a hybrid underwater vehicle manipulator (HUVM) system for cleaning of bio-foulings on a submerged pipeline in a sea. Features Image Processing through YOLOv9 architecture and Motion Planning through Bi-DRRRT* algorithm (Path planning, Obstacle avoidance, Slope-aware based). Developed at MARS LAB, Dept. of EIE, NIT Silchar.
Bio-Inspired Flagellate Based Motion of Underwater Robot
Design and development of an underwater robot mimicking flagellate motion for efficient, low-disturbance aquatic locomotion. Developed at National Institute of Technology Silchar, Assam.
Published Projects
Soft Actor-Critic Based Adaptive PID Control for Energy-Efficient Legged Robot Locomotion
Formulates legged robot locomotion as a Markov Decision Process (MDP), employing a Soft Actor-Critic (SAC) agent to dynamically tune PID gains in real-time. Simulated on a quadruped robot, the framework achieves a 20-25% reduction in power consumption without compromising trajectory tracking accuracy or gait stability.
Advances in Large Language and Vision Models for Robotic Manipulation
A comprehensive review synthesizing the transformative impact of Large Language Models (LLMs) and Large Vision Models (LVMs) on robotic control systems. Evaluates integration strategies such as multimodal fusion and hierarchical reinforcement learning, while critically addressing current research challenges.
Classification of Bearing Faults from Vibration Data of Induction Motor
A data-driven study assessing the effectiveness of One-Class Support Vector Machines (SVM) in detecting anomalies and bearing faults from the vibration data of induction motors. Utilizes Grid Search for hyperparameter tuning to accurately classify motor health conditions.
Design of a Robot Rover with 5 DOF Manipulator for Uneven Terrain
Research detailing the mechanical and system-level design of a highly capable robotic rover equipped with a 5 Degrees of Freedom (DOF) manipulator. Engineered to adapt to unstructured environments and perform complex physical interactions on uneven terrain.