Increased reuse levels through sorting of Second Hand using AI

Important results from the project

The purpose of the project is to support SRK´s continued development as an Second Hand-aktör and contribute to increasing sales in the stores, and increasing the recycling rate of goods. The goal of the project is to develop a system that automatically categorizes, weighs and captures the material of donations and where the system is taught to suggest whether the donation should be sold in a store or for recycling. An additional project goal is increased knowledge about the donations that are handed in to our stores, in terms of quantity and material content.

Expected long term effects

Within the project, we have developed an AI model that categorizes gifts and suggests whether the donation should be sold in a store or not. Thanks to the system, we now have a database with pictures and information about donated goods. These statistics are to be used to increase the proportion of material that will be sold for recycling and thereby provide increased income for Röda Korset. The knowledge we collected during the project, regarding system support, data on collected gifts and evaluation of work processes will be part of a decision basis for the sorting facility.

Approach and implementation

Three parties have been involved in the project; Chalmers Industriteknik, Svenska Röda Korset and Örebro Rödakorskrets. The project started up with a common process review and scoping of categories. As a next step Chalmers Industriteknik started the development of the image recognition program and AI model. At Örebro Röda Kors´ warehouse, the scanning equipment was installed. The training of the AI model was made by an operator photographing the donations and entering agreed information, in the system. The model was trained several times and became increasingly well-functioning.