Analytics toolbox and feedback (2)

D65 Analytics toolbox and feedback 2 41T61 Analytics toolbox and feedback CU MSC SIT EOS IVF The aim of this task is to develop standard feedback loops and analytics toolbox specified in WP1 It will enable optimization of the pilot lines through the analysis of the large amount data collected Depending on the type of data available T54 various types of process modelling approaches could be used to extract knowledge and features State of the art modelling data mining and machine learning tools will be reviewed eg Image processing techniques Support Vector Machine Principal Component Analysis Deep Neural Network data regression classification clustering and the most relevant will be tested with real data collected at different stages of the pilot lines Using the selected data analytics tools a wide range of information will be extracted and used to model various aspect of the AM process eg temperature profiles dynamic spatiotemporal imaging of the melting process electron optical or laser optical imaging but again not all information will be usable It is necessary to identify what standard outputs and related controllable factors could realistically be used automatically or manually at different stages of the pilot lines to enable appropriate and usable feedback eg tuning of the digital twin Manufacturing sequence adaptation or localized process parameter modification such as laser power Depending on the monitored information and on the controllable factors feedback loops will be designed using the online monitoring information in combination with data collected and models built from past machining experience These selflearning models could then be used to propose both onlineand offline changes to machining parameters and to manufacturing sequences depending on the controllable factors or to stop the building of components identified as faulty The main result of this task will be a range standard feedback loops demonstrating the reliability of the feedback process MSC will support CU analytics effort with Data Management and Data Quality monitoring in coordination with T14 T33 and T54 activity