Easy AI for Industry

The Coock project EARN (Eenvoudige AI vooR de iNdustrie), led by Flanders Make, focuses on making deep learning and artificial intelligence for computer vision accessible in production environments. Although machine learning has been used for image interpretation for decades, deep learning has recently caused a revolution. The technology sometimes surpasses human capabilities but remains challenging to access, especially for SMEs. Companies see the opportunities, but the entry barrier is high due to the technical knowledge required to set up and train neural networks. 

Flanders Make aims to bridge this gap by developing a software toolbox for defect detection, object recognition, object counting, and pose estimation. This toolbox will include collected algorithms and AI models from previous projects and will be accessible to end-users, technology developers, and integrators. Additionally, the deep learning framework will be set up in a cloud environment, and innovative methods for data annotation and training will be demonstrated. 

Another important aspect of the project is the development of PC and smartphone apps that use trained models to simplify the implementation and use of AI solutions on the shop floor. These apps will enable employees to analyze images and receive interpretations directly through their devices, thereby improving the efficiency and accuracy of production and inspection processes. 

The project also includes knowledge transfer and dissemination through workshops, events, publications, and educational videos. This will enable companies to evaluate AI and deep learning with low risk and investment. In summary, the EARN project aims to make AI and deep learning for computer vision accessible to industrial companies by offering tools, guidelines, and demonstrations, and broadly disseminating knowledge.