The SHIELD project (Long-term Structural Health monItoring of nuclear facilities using embedded acoustic Emission and machine Learning for Degradation detection) aims at adapting acoustic emission techniques for the monitoring of materials and components in nuclear installations, in particular when concrete is involved. The whole acquisition and analysis chain will be considered, starting with the study of suited sensors to embed in this harsh environment, specific denoising and triggering strategies to alleviate the noisy environments and machine learning methods to identify the nature of the events. In particular, optical fiber sensors will be considered, with their ability to monitor other parameters such as temperature or deformation. These developments will lead to a multi-parametric structural and materials health monitoring system.