Recycling old electrical appliances with AI

Household appliances have to be refurbished or recycled at the end of their service life. The Kikerp project aims to use artificial intelligence to collect information about old appliances and determine parameters such as model and condition in order to decide whether an appliance should be refurbished or recycled.

In order to support the sustainable handling of old electrical appliances in the sense of a genuine circular economy and reduce waste, researchers at the Fraunhofer Institute for Production Systems and Design Technology IPK are developing a multimodal AI assistance system embedded in a cloud-based architecture in collaboration with Yes Ecosystems Technology and HaKiGo. The AI is fed with data via a dialog-based application until the user has a classification for the proper processing, reuse and recycling of the electrical appliance under investigation. Both employees of electrical appliance manufacturers and end users are addressed.

Visual evaluation

The dialog-based front end runs on mobile devices such as smartphones and tablets. The application focuses on image-based recognition of products and uses pre-trained AI models to extract visual features and details. First, the employee uses the mobile app to record dialog-based information such as brand, product type, color and article number and then photographs the household appliance to record any defects such as scratches. On this basis, the AI visually assesses the quality of the goods, from which further measures can be derived, but also parameters such as price, condition, etc. can be determined. "Our AI modules are combined in a single architecture and run on a cloud server. They are operated via an interface on mobile devices," explains Vivek Chavan, a scientist at Fraunhofer IPK. The researchers use manufacturer data to develop the AI, but also generate artificial training data. They are also investigating whether an optical inspection can be trained using synthetic data. The aim is to test and qualify more than 5,000 household appliances with the help of AI by the end of the project and achieve a recognition rate of over 97%.

Fraunhofer Institute for Production Systems and Design Technology IPK, www.ipk.fraunhofer.de