Logistical performance measurement of photovoltaic manufacturing lines

The capacities of photovoltaic manufacturing lines grow steadily as well as the demand for photovoltaic products. Along with this growth, detailed performance measuring is desirable to control the factory as well as to find needs for optimisation. For different functions in a factory, different views are necessary. Line engineers have a different need for information than factory managers. Therefore different views in the form of figures, graphs and statistics are needed, which have to be based on the same consistent factory data. In a typical factory, large amounts of data are collected. To draw conclusions and give recommendations, this data most often has to be condensed into specific statistics. Fraunhofer IPA has developed a method and a tool, which allows the generation of specific statistics from the viewpoint of factory logistics. The basis of this is a log file, which tracks the movement of all entities in the factory. For instance in a thin film manufacturing line each substrate can be tracked, in a wafer based manufacturing line a lot could be the tracked entity. Each entry in the log file describes a specific event (e.g. arrival of a substrate, start of processing, removal of a substrate, scrapping of a substrate). These events are logged in a timely manner. For each entity in the manufacturing line a detailed record of all interesting events is gathered to allow seamless tracking throughout the whole lifetime in the factory. However, aggregated views are necessary for performance measuring, which typically describes performance indicators like cycle time, waiting time, down times, etc. for a specific period of time. The method which is now applied is the aggregation of the events in the log file into statistics that delivers indicators like average cycle time, waiting times of substrates in front of specific process equipment, average buffer utilization, etc. Inside the tool a state machine is utilized. This state machine can be graphically created and edited. In this state machine specific events of a substrate are gathered and accumulated to a single data set. For instance a data set could consist of the arrival of substrate at a process equipment, start and end of processing and retrieval of the substrate. The results of the state machine are fed into a database which allows queries to aggregate the data.

Investitionen in große Photovoltaik-Anlagen können zu einem hohen Kapitalaufwand führen. Um eine schnelle Rendite zu erzielen ist es unbedingt notwendig, eine hohe Nutzung der Prozessausrüstung zu gewährleisten. Die Leistungsoptimierung einer Photovoltaik-Anlage erfordert ein fundamentales Verständnis der Prozesse sowie des Materialflusses und der Produktionsausrüstung. Das Fraunhofer IPA hat eine Methode zur Gewinnung und Auswertung von Betriebsdaten entwickelt um die logistischen Einflussfaktoren zu erfassen und zu verstehen. Mit diesen Betriebsdaten kann die Leistung von Materialflusssystemen und Produktionsausrüstungen ausgewertet werden, was die Entdeckung und Eliminierung von Ineffizienzen in den Produktionslinien ermöglicht. Die in diesem Artikel behandelten Methoden zur Beschaffung und Analysierung von Daten zur Betriebsleistung beziehen sich hauptsächlich auf Dünnfilm-Produktionslinien, sind jedoch auch für kristalline Technologien anwendbar.