"Laser welding is a widely used technique in automotive, aeronautical and shipbuilding industries. New developments are being investigated and it is expected that in the near future it will replace traditional welding techniques in many applications. During laser welding a free electron plasma is generated. The radiation emitted by that plasma and the plasma itself can be characterised by means of several specific measurements, among which specific wavelength radiation intensity, plasma electronic temperature and plasma electronic density are featured. Laser welding process is parameterised according to the characteristics of the specimens to be welded. Each type of specimen has its own parameter set, found during a design stage prior to fabrication. But due to laser facility degradation, ageing or unknown parameters that can influence the welding results the initial parameter set can loose its suitability, which induce human operators to tuning themselves the facility according to their knowledge. Moreover, manufacturing new things may require the calculation of a new parameter set, which is a time-consuming, waste-producer and cost ineffective process. Besides this, the problem of defective welded seams is always present. Defects cause product quality degradation and cost increase. Sometimes faults are originated by an inappropriate specimen preparation, but often an inadequate laser process tuning causes them. The objective of this project is to provide laser-welding facilities with an adaptronic module that, using knowledge-based intelligent algorithms founded on on-line self-optimisation and learning, gives the facilities self-adaptive capabilities for coping with changing production environments, allowing rapid reconfiguration and providing defect avoidance capabilities. This will be achieved by means of multi-layer control, plasma and radiation related sensing and actuator structures."