DIPANSH SINGH

Conducted root cause analysis of welding defects for multiple vendors of Bajaj Auto, utilizing data cleaning, data preprocessing techniques in Python. Employed libraries such as pandas, seaborn, and matplotlib, along with processing methods like Fast Fourier Transform and curve similarity measures. Developed algorithms to identify anomalies in a dataset of 1.2 million data points. Designed Power BI dashboard for project monitoring. Leveraged Microsoft Power Apps to develop user interfaces for manual department activities, later automating them with Power Automate. Connected Excel data to Power Apps using Excel VBA scripts. Engineered time-saving algorithms in VSCode to automate manual costing for manufacturing components using a 3-D modeling soft- ware API. Enhanced productivity by reducing costing time from 15 days to just 30 minutes. Implemented automated cost sheet generation through Excel scripting.

Developed a paint defect inspection system for assembly line using machine vision. Built a model with image pre-processing techniques like edge detection, image morphing, image masking and bitwise operation using open-cv and deep learning classification model using Resnet with accuracy of 94 %.

Improved tool life through machining data extraction using SQL from KEP Server, followed by data cleaning, analysis and visualization in Python. Identified trends in tool life cycle, analyzing correlations with over 10 parameters, leading to 33 % increase in tool life. Created a centralized database containing weights of all parts in Bajaj models and integrated it with billing platforms for automatic updates. Devised a parameter-based approach for conducting comparative analysis among various 3-D scanning products and negotiated with vendors for their procurement.