Keep on running: the power of predictive maintenance

Digital printing generally involves lots of short run jobs with fast turnaround for ever-shorter delivery times, all largely driven by automated workflows that might span several different departments from prepress and printing to finishing and fulfilment. Unexpected breakdowns cause trouble with close production schedules, particularly where printers have relied on one or two high volume devices. To counter these problems and reassure customers, many printer vendors are embracing predictive maintenance with the promise that it will reduce machine downtimes.

Inevitably, machines break down and it’s impossible to eliminate such problems completely, but reducing the frequency of such problems would take a lot of pressure off of most print operations. In the past, printers have relied on set maintenance routines that would take account of the expected service life of major components. However, this sort of preventative maintenance is a very inexact science as many factors, from ambient conditions, the workloads involved and the operator’s skill in using the printer will all affect each machine and its parts differently.

That said, modern printers now bristle with an array of different sensors to record a huge range of information, such as the working temperature inside the printer, as well as the amount of time the printer is on and running. This data can give a very accurate idea of the deteriorationon the printer and the level of work that the system overall is doing. Most printers also include specific sensors to measure the performance of all the major components to better anticipate when those components might fail.

Of course, this raw data only gets you so far. You still have to be able to access it and to analyse it. To that end many equipment vendors now sell analytics programs alongside their machines that can interpret this data and help customers optimise the efficiency of their system. These programs have allowed the machine builders to gather data on all the systems they have in the field so that they can better understand how varying criteria between different customers can affect how their equipment performs.

Naturally, this does produce an enormous amount of data but this is one area that has really benefited from artificial intelligence and its ability to crunch through huge datasets to find patterns and predict how those patterns might work in real world scenarios. As a result, this type of software is increasingly able to give…