Projects
Research & Digital Twins
Digital Twin-Based Optimization of Biomining
Research on sustainable Rare Earth Element production. Introduced a refined method for phylogeny estimation significantly reducing computational complexity.
Graph-based Digital Twins for Supply Chain
A theoretical framework for management and optimization of complex relational systems across manufacturing, transportation, and supply chain networks.
AI-Powered Smart Manufacturing Agent
Researching the integration of Digital Twins with Large Language Models (LLMs) and AI agents to enable intelligent automation and decision-making.
Industrial Engineering
Process Flow Optimization in Pre-Assembly Line
Walton Refrigerator Plant-3
Analyzed production flows using VSM & SMV. Consolidated 4 production lines into one JIT line, cutting buffer inventory and workforce by 20% while increasing overall productivity.
Design of Overhead Monorail System
Walton Refrigerator Plant-3
Designed a Slat Conveyor and an Overhead Monorail System for Side Cabinets. Conducted cost analysis (617,500 BDT) and calculated a break-even period of 12 months.
Machine Learning & NLP
Traffic Diffusion World Model
Developed a world model for realistic traffic simulation using Diffusion Probabilistic Models. This framework predicts complex traffic dynamics and agent behaviors to enhance autonomous driving planning and decision-making.
View RepositoryAutomated Essay Scoring System
Developed a model to automatically score essays by evaluating content relevance, grammar, and coherence. Utilized NLP techniques like word embeddings and deep learning architectures (LSTM, BERT).
Credit Scoring Predictive Model
Built a model to predict individual creditworthiness using past financial data. Applied classification algorithms like Logistic Regression and Decision Trees to assess risk.
Emotion Recognition from Speech
Created a system to recognize human emotions (happy, angry, sad) from speech audio by applying speech signal processing and deep learning techniques.