Projects Portfolio

Industry 4.0 Readiness Platform

Made by Abe, Saii, and Sammie for the Internship-CSI program.

Data-readiness software and strategic roadmaps to guide manufacturers in their transition towards digital transformation and Industry 4.0 solutions.

What I do:

  • Conduct industrial on-site audits to collect manufacturing data for Microsoft's Co-Innovation Lab at UWM.
  • Perform AI readiness assessments and develop digitalization roadmaps to help manufacturers adopt Industry 4.0 solutions.

Project Features:

  • Questionnaire
  • Dashboard
  • Learn about Digitilization
  • Case Studies

Project Progress:

Planning & Questionnaire 80%
Start Questionnaire View Dashboard Ask our AI Agent

AI-Driven Gear Fault Diagnosis

Academic Research Project | Jul 2021

Developed an advanced gear fault classification system using neural networks and deep learning. The system leverages signal processing and wavelet-based feature extraction from vibration data to accurately identify and diagnose faults in industrial machinery, enabling predictive maintenance.

Methodology:

  • Analyzed vibration data from a gearbox test rig with 6 distinct gear health states.
  • Engineered features using Maximal Overlap Discrete Wavelet Packet Transform (MODWPT) and entropy.
  • Trained a Multi-Layer Perceptron (MLP) neural network for classification.
  • Implemented a 1D Convolutional Neural Network (CNN) for end-to-end deep learning.

Key Achievements:

  • Achieved classification accuracy of 99.3% with the CNN model.
  • Successfully identified faults under 15 different operating conditions (speed and load).
  • Demonstrated the superiority of MODWPT over traditional LMD for feature extraction.
  • Developed a user-friendly visual tool to highlight the detected fault type on a diagram.
TensorFlow Keras Scikit-learn Signal Processing