Three common misconceptions of prognostics in manufacturing

03 March 2016

Prognostics – predicting when a machine will stop being able to perform its given function (protecting machine uptime) has been around for a while but some old beliefs have held it back from really taking off. With the rise of the Industrial IoT and new technology though, some of the ‘facts’ about it are now misconceptions:

 

It’s expensive

Typical condition monitoring and prognostics projects for forecasting machine failure get their major expenses in two ways: the time needed to develop bespoke models of the physical assets being monitored and the assets that need to be run to destruction to understand their failure modes.

This is the old way of doing things. Thankfully with Senseye, we don’t rely on building up these bespoke models and we don’t need to rely on running your assets to destruction – those costs are completely avoided; making it an affordable solution for every manufacturer.

 

You need a great amount of existing information

A typical prognostics project would need a lot of existing data, perhaps several years’ worth. Not with Senseye – since we can process machine failure information in the cloud for known fundamental components (bearings, motors, actuators etc.), we know what these failure modes look like. When combined with realtime information from your machinery, Senseye can start providing valuable information without needing a historical backlog – ideal for retrofits. Although of course if you already have this information in a data historian, it can only help!

 

It’s a manual process

Current solutions require manual interpretation of raw data, this could be many gigabytes per day of raw sensor values, requiring an expert to know what they’re looking for. Data scientists and engineers are not cheap and relying on manual interpretation is also relying on that individual not being tired, grumpy or otherwise distracted.

 

Since Senseye is automated, it crawls over all of the raw data, extracting and interpreting key features automatically and in a fraction of the time, producing machine failure forecasts without any effort. Of course the raw data can still be manually inspected but it’s comforting to know that the process just got a lot easier.

 

Tell me more

We’re working with world-leading companies to help them embrace the advantages that Senseye can bring to machine failure forecasting. Book a demo to learn more!

Request a demo of Senseye

prognosys_lights.jpg

Keep me notified

Most Read

Try Our Free ROI Calculator