For some years press manufacturer manroland Goss has been collecting data on how their presses are running, with the idea of telling users what's right and wrong with them.

An interesting session on the second day of the virtual World Printers Summit delivered an update on that project, and its potential for the future.

Technology and development vice president Manuel Kosok (pictured) introduced delegates to the acronyms at the heart of the project: the familiar TCO for total cost of ownership, and OEE for overall equity and effectiveness (or GAE if you speak German).

The huge development project is focused not only on reducing costs when the press is running, but also when it's not running, and with the immediacy of a Zoom connection, 41 per cent of delegates responded that they knew their OEE, and another 23 per cent "knew partially".

manroland Goss's Maintellisense processes machine availability data which, Kosok says, "most technical employees don't have time to process data or deal systematically with the analysis".

Data gathered over ten weeks was used as an example

Using data gathered over ten weeks from a newspaper press and two commercial presses, the system charted the status of each press, the failures that occurred, and solutions for the problems identified.

In one case, spotting a shortage of oil in bearings potentially saved two-to-five days' downtime. In another, spotting trends in gross and net waste pointed to web breaks - ten of them a week - typically following splices and, when that problem had been fixed, moving to just after 'impression on'.

An analysis of waste helps identify problems

Fixing those problems is saving the user an estimated 100,000 Euros a year.

Kosok says with users printing 174 trillion copies a year, "we see a lot more things to do" with the technology, including extend it to non-manrolandGoss users. Future projects include predictive maintenance, the use of AI and peer group data, and - through register control data - what it's taking to produce a quality product. First applications of AI, despite the complexity of press data, are likely to include web tension and variation.

The issue of which operators are performing best is a sensitive one, given legislation in some countries, but yes, "indirectly it can be seen".

Peter Coleman