Is it?

Your data shows when your machines are down.
Look close and data will also tell you how to avoid it.

Manufacturers and machine owners know that every minute that a machine is not running costs money in missed production, lost revenue, lost customer confidence, and possibly lost future business. So machine makers and owners are working hard to make their machines as dependable as possible. Rigid, redundant designs provide for long term reliability. Operating procedures ensure safe and efficient operations. Regular maintenance ensures the grease stays on where it needs to. Yet, invariably the unexpected happens: parts fail, go out of alignment, or quality issues force unplanned downtime, despite best maintenance practices.

Many of such failures can be predicted, and their impact therefore avoided by carefully watching machine parameters. For example, failing motors or gears develop unusual current draws or vibrations long before going kaput. Let us show you how to make Data Science a reality by tackling the most frequent downtime reason experienced by your machine. Our machine learning algorithms can detect data patterns that are appearing repeatedly prior to the most frequent downtime events in your data history. We detect and isolate these data patterns, and continuously look for similar patterns in real-time data. Once spotted, an alarm is raised to alert operators of impending failures, before they happen.

In many cases, no special sensor data or equipment is needed. The normal IoT data which you are already collecting is holding ample clues to hint at failing parts. Combined with engineering knowledge about your machines, our data science team can go even further by fine-tuning our algorithms to provide detailed insights into the future health of your machines, and the impact of any predicted failure on your production, and bottom line.

Our algorithms can detect patterns across as many variables as your machine creates – maybe hundreds – making this detection algorithm much more capable than any human or manual “spreadsheet analysis”. We have tools to create this algorithm quickly and using our proven methodology, allowing you to reap the benefits of data science with low risk.

Sabel - andon display

An example dashboard showing the OEE details and trends, plus the prediction of estimated life remaining of a machine part.

Get in touch with our Data Science team to talk about how we can help you avoid unplanned downtime.  Click here.