Big Data Analytics: A Transparent View of Banknotes
Banknotes can be produced more efficiently if the rate of printing faults is reduced. Single note inspection alone is not enough, but with smart analysis software, printing plants can use their data to limit the number of faults in the process and resolve any faults early – or at least to learn from issues and achieve a fact-based continuous improvement process.
Ten million print sheets with 54 banknotes each – as with the second series of 10 Euro notes – means 540 million potential sources of faults which have to be detected in single note inspection. After cutting, at least 100 relevant properties are checked for each banknote. Not only does this require outstanding technical capabilities, it also drives up the volume of data generated.
Previously, it was virtually impossible to use quality control data as feedback in the printing and substrate production processes. Single note inspection systems are simply not able to record banknote metadata, such as input unit or sheet position. Moreover, it’s impossible to aggregate the results of multiple banknote processing systems (BPS) or print batches. That’s why data visualization tools like G+D Currency Technology’s “BPS Eco-Analyzer” offer “snapshots” – for example, statistics for each 50,000 banknotes, or the development of individual data spots over time.
“It’s much more interesting when you analyze all the data from a print job,” says Dr. Marcus Schmeißer, Senior Big Data Product Manager at G+D Currency Technology. With the Banknote Quality Analytics tool “SeriTrack LUNA”, all the data from multiple BPS X9 systems can be merged and analyzed during single note inspection.
Banknote Production: Reduce Costs and the Reject Rate
BPS X9 and SeriTrack LUNA
All printing plants want to reduce costs and keep the reject rate as low as possible. The combination of efficient hardware and advanced software helps optimize the cost efficiency of their process. Data from BPS X9 single note inspection systems can be merged in SeriTrack LUNA, a banknote quality analysis software. To take one example, the tool uses a banknote’s serial number to calculate its position on the print sheet and which print unit it belongs to. This allows the printing plant to identify steps in the process that did not run smoothly, and reasons why the banknote was finally classified as waste.
Whether it’s a weak printed image, too much or too little pigment, scuff, register problems, or flaws in the substrate – it’s too late to fix notes that have already been printed. “A small print batch allows you to learn lessons for the next job,” says Dirk Fischer, Finishing/Printing Plant Systems Expert at G+D Currency Technology. “But with a big print job, you can still take action before the job’s complete.” Traceability, which is already a major concern in the pharmaceutical industry, is another issue. Demand for traceability is also on the rise in banknote printing, says Fischer: “More and more clients want the whole chain to be perfectly traceable, from the paper roll to the individual note.”
From raw data to comprehensive analyses
In a BPS X9, each banknote must travel around two meters to reach the reject gate, where its fate is decided: stack or shred. A highly complex sensor system uses infrared, UV, magnetism or ultrasound to inspect the banknotes and delivers raw data, which is processed by computers and compared with threshold values. “It takes 22 milliseconds to decide whether this banknote meets all the requirements,” says G+D manager Dr. Schmeißer. A BPS X9 can check 44 banknotes per second.
Real-time evaluation takes place within the system itself. However, measurement results are stored in a data base. From there they can be sent to the SeriTrack LUNA tool, which is located in the G+D Currency Technology data center. In a normal print job, this amounts to between 40 and 200 million data sets. These are assigned to the appropriate production unit and combined into incidences, explains G+D expert Fischer: “For example, the analytics engine might detect in one print sheet that the pigment is repeatedly smeared in the area of the blue European flag. With this information, the person responsible for printing can then prioritize and specifically search for causes and/or solutions for faults.”
Analytics as a Service for Printing Plants
To begin with, G+D Currency Technology used SeriTrack LUNA exclusively in-house, to optimize its own printing processes. “Now we also offer this tool to other printing plants from our data center, in the form of Software as a Service (SaaS),” reports Big Data Manager Dr. Schmeißer. Special consulting services help clients to eliminate printing errors and, at the first stage, to learn how to operate the application. “We are currently working on also developing the software as an on-premise installation for client data centers.”
More and more clients want the whole chain to be perfectly traceable, from the paper roll to the individual note.«
Big Data Analytics allows a Transparent View of Banknotes
In future, data analysis won’t just be limited to optimizing print processes, but will extend to the full banknote lifecycle. “By combining results from the printing plant and from circulation, we can find out more about the behavior of banknotes and optimize them further,” says finishing expert Fischer. Ultimately, the notes can be identified by their serial number and their data sent to analytics software.
As well as horizontal integration within a printing plant, vertical integration between the paper mill, printing plant, and central bank means a further step towards Industry 4.0 – and ultimately towards data transparency that allows comprehensive improvement of efficiency. For example, if the printing plant makes the aggregated quality results of the banknotes supplied available to the central bank, then incoming product inspection can be made considerably simpler. Additionally, if the central bank also receives the actual single note inspection data, it is possible to identify changes to machine-readable properties of banknotes in circulation, or increases in contamination. This information allows the cash cycle to be further optimized, since logistics concepts can be customized, or optimal security features for future series of durable banknotes can be extrapolated on this basis. In this way, Big Data analytics, and the decisions taken on this basis, support one of the strongest pillars of the entire cash cycle: mutual trust between stakeholders.