Independant Robotics Researcher specializing in AI, Robotics, Embedded Systems, Vision Systems, Intelligent Control & Data Engineering
| Category | Technologies |
|---|---|
| Languages | |
| Embedded & PLC | |
| ML & Computer Vision | |
| Robotics & Control | |
| Design & Tools |
1. Autonomous Indoor Robot with Adaptive Decision-Making • [Ongoing]
Mobile robot implementing following-in-front navigation paradigm with real-time simultaneous localization and mapping (SLAM). Uses predictive control algorithms with interactive decision-making strategy to resolve path ambiguity at T-junctions and cross-intersections through sensor fusion and human-robot interaction protocols.
ROS2 SLAM Toolbox Python Raspberry Pi IMU Sensor Camera
Adaptive Neuro-Fuzzy Inference System (ANFIS) controller for magnetic levitation platform. Combines fuzzy logic reasoning with neural network learning to handle nonlinear dynamics and achieve stable position control with minimal overshoot and steady-state error.
MATLAB ANFIS Fuzzy Logic Control Systems
Integrated Long Short-Term Memory (LSTM) network with ANFIS controller for time-series prediction in nonlinear magnetic levitation systems. LSTM captures temporal dependencies while ANFIS handles real-time control, enabling improved response characteristics compared to standalone approaches.
MATLAB LSTM ANFIS Neural Networks
Convolutional Neural Network (CNN) trained on Extended MNIST (EMNIST) balanced dataset for multi-class character classification. Architecture includes convolutional layers, max pooling, dropout regularization, and dense layers to classify 47 character classes (uppercase, lowercase, digits).
TensorFlow Keras OpenCV Python
Closed-loop PID control system for aerodynamic ball levitation using ultrasonic distance feedback. Implements system identification techniques in MATLAB to derive transfer function model and optimize controller gains through iterative tuning methods.
Arduino MATLAB Python Control Theory
Multi-floor elevator control system programmed in 8051 assembly language. Features interrupt-driven architecture with timer-based error detection, load cell integration for weight sensing, limit switches for floor detection, and PWM motor control for variable speed operation.
8051 Assembly Proteus Embedded Systems