Quick Summary
- ▸ CodeBranch's dedicated development team worked on this project, which was an AI image recognition system that extracted the dimensions of a metal plant's waste and stored them in an inventory for use in future projects.
- Delivered a system based on genetic algorithms that recognises the shape and dimensions of scrap metal pieces.
- Enabled the metal company to optimize resources and save money by systematically reusing waste material in future projects.
Overview
The system was developed for a sheet metal plant. The main objective was to create a system that catalogs the inventory of waste based on shapes, dimensions, gauges, and materials. To create the inventory, the waste was placed on a table to be photographed. Using an AI image recognition system, the dimensions were extracted and the waste was stored in the inventory system. Each time a new project was started, the system — based on genetic algorithms — checked if there were pieces in the waste inventory that met the specific needs of different parts of the products to help the recycling process. In addition, if a new sheet was to be used, the system suggested how the sheet should be cut to reduce waste as much as possible.
Industries
Services Provided
Approach
The technology used in this project included PolyK for polygon shape computation, Genetics-JS for running genetic optimization algorithms, JavaScript for application and UI logic, and Python for AI image recognition and backend services. Waste pieces were photographed on a designated table and processed by the AI model to extract dimensional data automatically. The genetic algorithm then handled the matching and cutting optimization logic. This project involved 4 developers (2 senior and 2 semi-senior), 1 QA expert, and 1 UI/UX designer.
Results
- Delivered a system based on genetic algorithms that recognises the shape and dimensions of scrap metal pieces.
- Enabled the metal company to optimize resources and save money by systematically reusing waste material in future projects.
- Reduced raw material costs by matching existing scrap inventory to new project requirements before ordering new stock.
- Minimized future waste generation by providing AI-driven cut optimization suggestions for new metal sheets.
