ABSTRACT

This project aims to develop a mobile application that enhances the accuracy of building cost and material estimation, addressing the discrepancies often encountered between estimated and actual quantities. By improving cost management and decision-making in construction projects, the application provides a valuable tool for industry professionals. The frontend of the application is built using the Flutter framework, offering users an intuitive interface to capture essential parameters required for precise cost estimation. The primary objective is to streamline the estimation process and mitigate complexities associated with managing a separate database, ultimately facilitating better project planning and resource allocation.

The adoption of an iterative and incremental development methodology was used, ensuring continuous refinement and responsiveness throughout the development process. It consists of two main components: the frontend and the backend. The frontend, built with Flutter, provides a user-friendly interface for inputting necessary parameters. On the backend, the entire codebase is hosted on the Render platform, eliminating the need for a traditional database. Instead, a pickle file located in the same directory as the main code is utilized, simplifying the estimation process and enhancing efficiency. The model employed in the project utilizes the Random Forest Regressor algorithm and achieves an accuracy score of 74.3%. To further improve accuracy, future iterations could involve expanding the training dataset, refining feature selection and engineering techniques, fine-tuning model hyperparameters, and exploring advanced machine learning algorithms.

The project's results demonstrate that the developed mobile application successfully enhances the precision of estimating building costs. With an accuracy score of 74.3%, the application effectively reduces the discrepancies between estimated and actual quantities. The correlation between the input parameters captured by the frontend and the predicted costs is significant, validating the reliability and usefulness of the developed solution. The iterative and incremental development approach allows for continuous improvements, and future enhancements can be made by incorporating more data, optimizing feature selection, and exploring advanced machine learning techniques. Overall, this project provides a robust and efficient mobile application that empowers construction professionals with accurate cost estimation, leading to improved cost management and decision-making in construction projects.