There are three tasks that are being researched as part of WP3
Task 3.1
Will focus on building and validating a laboratory prototype MPI optical sensor(s) for measuring film critical elements, as informed by, and meeting criteria specified by WP1 and WP2. Key challenges will be auto focussing the instrument, overcoming substrate vertical movement, sensor calibration and size/access constraints within the manufacturing scenario. Our project partners (CPI, Emerson & Renwick, FOM Technologies) will assist in defining these physical constraints via real (or virtual / CAD) instrument access.
Task 3.2
Will look to develop a range of low cost, easily and widely implemented “proxy sensors” based on either darkfield imaging which can do large area coverage or will be light-based non-imaging sensors. These proxy sensors offer low data quality thresholds, yet allow full film width analysis (approaching metres wide) achievable via multi-sensor installations. These sensors allow indirect metrology data (i.e. opaqueness) at incredibly fast data rates, albeit data-rich, but direct information poor, all of which can be fed into ML (WP1).
Task 3.3
Will focus on accelerating the data handling and computing process using graphic processing to develop metrics such as a running defect maps and continuous (low data density) film metrology data and will work closely with WP1 to develop data communication protocols and defects classification. Through interactions between WPs we hope to identify which sensors (MPI/Proxy) and their locations within the drying process (equivalent to time information) generate the most useful data for the ML process. WP3 will also transfer prototype sensors to WP4 for lab/in-house acquisition of data and transfer site scale data to expedite the development of WP1, based on parameters set within WP2.