Design and Analysis of Intelligent Control Systems in Minimally Invasive Surgical Robotics for Enhanced Procedural Accuracy:
A simplified but clinically relevant 2-DOF surgical manipulator model was constructed, incorporating forward kinematics, approximate dynamics, measurement noise, and external disturbance simulation. A hybrid intelligent control architecture was developed, integrating a baseline PD controller with an adaptive parameter estimator and a supervisory safety layer. The full system was implemented within a modular Python simulation framework, enabling iterative experimentation, data collection, and visual analysis. More than thirty-five simulation experiments were conducted across varying test conditions, including sinusoidal tracking, step-response behavior, Gaussian noise injection, drift scenarios, and disturbance-rich environments. The intelligent control system consistently outperformed classical PD control, achieving lower RMS tracking errors, faster disturbance recovery, and superior robustness. Adaptive compensation effectively counteracted dynamic uncertainties and sensor drift, while the supervisory layer maintained stability by regulating torque spikes and smoothing noisy signals. End-effector trajectory deviations were maintained within sub-millimeter ranges, satisfying typical MIS precision requirements.
DNA as a Data Storage Device:
In this project, DNA was explored as an innovative medium for high-density data storage. The work focused on evaluating the feasibility of encoding and preserving digital information within synthetic DNA sequences. The project involved assessing DNA’s long-term stability and its capacity to store large volumes of data compared to traditional storage technologies. Various encoding and decoding methods were studied to determine optimal strategies for converting digital files into nucleotide sequences while minimizing errors during synthesis and sequencing.
Additionally, multiple approaches for embedding data into DNA were examined, including different coding schemes, redundancy techniques, and error-correction mechanisms. Experiments were conducted to understand how environmental factors—such as temperature, humidity, and storage conditions—affect DNA integrity and data retention over extended periods. The project also explored the practical challenges associated with reading, writing, and retrieving information from DNA, providing insights into its future potential as an ultra-dense, durable, and biologically inspired data storage platform.
Automated Water Irrigation System:
This project involved developing an AI- and IoT-enabled automated irrigation system designed to optimize water usage and enhance crop management. The system continuously monitors soil parameters using a 7-in-1 soil sensor capable of measuring moisture, temperature, pH, conductivity, and other key indicators. Based on real-time soil moisture data, the IoT controller autonomously regulates water flow to ensure precise and efficient irrigation, reducing water wastage and improving crop health.
Beyond irrigation control, the system leverages machine learning models to analyze soil characteristics and predict the compatibility of various plants and crop rotation patterns for specific plots. This predictive capability helps farmers make informed decisions that support sustainable agriculture. The project also focused on optimizing sensor placement and water distribution levels to maximize measurement accuracy and irrigation effectiveness. Overall, the solution integrates hardware, AI analytics, and automation to provide a smart, resource-efficient approach to agricultural management.
