Reactive activities can also be avoided to some extent, by monitoring the process and predicting when an intervention is necessary. For electrical maintenance, consider equipment such as handheld thermal imaging cameras, which can be used to help monitor temperatures, and identify hot spots that can denote issues such as faulty insulation and electrical switch gear issues, and near capacity fuses. Components including sensors, valves and even PLC I/O can degrade over time, causing problems including process accuracy and stability. Monitor performance using process calibrators.
Similarly, to get advance warning of mechanical failures, portable vibration analysis equipment can be used to make repeatable, severity-scaled readings of overall vibration and bearing condition.
Excessive wear can be a primary cause of failure for gearboxes. Oil and grease analysis can be performed to understand lubrication condition, including contamination and debris content, allowing appropriate action to avoid failures and downtime.
These are all great steps to hone your processes - but embracing the Industrial Internet of Things, (IIoT) will allow you to really step things up.
Sensor technology can be the start of the predictive maintenance journey
Before getting in a spin about Industry 4.0 and the complexities of IIoT, there’s no need to try and run before you can walk. Start small, and consider where to deploy the right pieces of technology to best mitigate negative impact on your operations. Sensor technology has become much more accessible from a product price and implementation perspective, and provides a good ‘toe in the water’ approach. Identify where your data sets might need enhancing to best avoid failures – whether that be in the areas of vibration, electrical energy, temperature or pressure – and look to strengthen existing data sets with the use of sensors to gather further information in real-time. Focus only on collecting the data you need – big data is a well-used term, but by sticking at first to small data, gleaning actionable insights will be more achievable, which is critical in the predictive maintenance mission.
Setting parameters of what falls into normal or abnormal operation means that abnormal can be quickly flagged and used as the trigger for deeper investigation into a potential issue. Many organisations already have more data than they think they have, and this can be enhanced with frequent real-time data that can be gathered using sensors and fed into a simple recording method such as Excel or a more sophisticated analysis tool.