Structural Engineering
G + 9 Storied Residential Building (AutoCAD & Revit)
Designed a G+9 storied residential building using AutoCAD and Revit, featuring a ground-floor parking level and nine residential floors. Each floor accommodates two residential units, with a unit area of 1,358 sq ft each. The design prioritizes architectural comfort, spatial efficiency, natural ventilation, and functional planning, ensuring livable and user-friendly residential spaces.
Countryside Cafe Modeling (Revit & Twinmotion)
Designed a single storied cafe on an area of around 1,620 sq ft. The project focused on efficient spatial planning, functional zoning, and customer comfort, ensuring smooth circulation for both patrons and staff. The design considered flexible seating layouts, service counter placement, and back-of-house functionality, while maintaining clarity in architectural documentation through BIM-based modeling and coordinated drawings and 3D rendering using Twinmotion.
Machine Learning (Supervised)
Modified Concrete Strength Prediction
Developed a machine learning model to predict the compressive strength by preprocessing data, training multiple machine learning models to predict the compressive strength of a hybrid concrete type based on key input features. The best performing model random forest regression was further optimized to improve prediction accuracy. [Link]
Deep Learning (CNN)
Personal Protective Equipment Detection
The goal of the project is to develop a deep learning based model to detect whether the workers at construction sites are wearing proper PPE using YOLO pre-trained deep learning model and libraries like OpenCV, cvzone and math. The objects of interest include personal protective equipment (PPE) items like Gloves, Hardhats, Masks, and Safety Vests. [Link]
Surface Crack Segmentation and Categorization
Worked on establishing a computer vision model to automatically identify and segment cracks on surfaces to support structural safety and maintenance, and annotated concrete surface images are being used for pixel-level crack detection and analysis.
Crack Labeling
Crack Segmentation
Vehicle Detection and Counting on Road
Worked on to develop a model to count and track each individual vehicle in a video as it moves across a predefined region using YOLO model, SORT tracking algorithm, and OpenCV and cvzone libraries. [Link]
Data Analysis
Weather Dataset
Weather Dataset is a time-series data set with per-hour information about the weather conditions at a particular location. It records temperature, dew point temperature, relative Humidity. Wind Speed. Visibility, Pressure, and Conditions. This data is available as a CSV file, and has been analyzed using the Pandas Data-Frame. [Link]
Police Checking Dataset
The data from a Police Check Post is available as a CSV file. It demonstrates stop_date, stop_time, country_name, driver_gender, driver_age, driver_race, violation_types, search_conducted, stop_outcome, stop_duration, and drugs_related_stop. We are going to analyze this data set using the Pandas Data-Frame.[Link]
COVID-19 Dataset
A small dataset of Covid-19 has been chosen for analysis. The dataset used here is till 29th April 2020 and it has records of data, states, region, confirmed, deaths and recovered counts. This dataset has been analyzed using the Pandas Data-Frame. [Link]
Census Dataset
This dataset is about total population , demography , literacy , districts , states, workers , religion . education and age. The data used here is of 2011 India Census of each district. [Link]
