By Splunk January 16, 2019
With each industrial revolution comes technological advancements that enhance methods of production, all the while introducing new opportunities for failure. For example, the assembly line generated previously unseen production levels, but would ground the entire operation at the smallest problem. Redundancy in parts and labor—an expensive fix—was employed to address this.
Recently, the third industrial wave took the form of digital technology and automation, which again increased production and efficiency, but also brought with them their set of problems (e.g. issues of data integrity, complexity and security). The fourth wave, aka Industry 4.0, is addressing these issues with transformative technology including IoT, cloud, machine learning, analytics, and even augmented reality for industrial settings.
We’ve finally reached the point at which we no longer have to be reactive or preventative. Condition-based and predictive maintenance are starting to change the industrial world.
Moving Toward Predictive Maintenance
Today, the most common form of maintenance happens in preventative fashion. Much like we handle a car’s oil changes, there is an estimated interval for when it makes sense to look at industrial machinery and maintain its health. But considering cost, it would be amazing to stretch out that interval as long as possible before maintenance costs are incurred—without actually breaking anything.
This is where Industrial IoT comes into play. There are new levels of connectivity, machine learning-powered analytics, and the convergence of OT and IT technology—all of which is coming together in the cloud, providing a new level of monitoring and diagnostics for predictive maintenance.
The truth is, preventative maintenance misses most machine failures because it’s taking a time/schedule-based approach to a problem that is not time-based about 89 percent of the time—think about the initial break-in period or infant mortality of new machines and assets.
This can be costly, leading up to millions of dollars a day. One minute of downtime can cost up to $22,000 dollars in the automotive industry. The urgency behind unplanned maintenance can also raise the cost by 12-15 percent due to the urgency and unforeseen nature of it. There is also the effects of unplanned downtime besides true downtime cost (TDC), namely things like lower production levels, labor overhead and possible equipment loss and replacement.
As a result, forward-leaning organizations are looking at predictive maintenance strategies. It takes into account estimated service intervals, while also bolstering maintenance decisions with data-driven insights based on measurements of operating conditions. Looping back to the car, we can now also look at oil viscosity and engine speed, or even take into account external data points like temperature and location. Powered by statistical thresholding and forecasting, organizations can go predictive to better manage parts and labor costs, not to mention general asset availability.
Don’t Get Left Behind
There are challenges to predictive maintenance like the process of digital transformation and modernization. Operational technology stacks often lag behind those of IT. There’s also the issue of data velocity, variety and variability. There is so much data and data types in informational and operational silos, which often make it inaccessible—but these are hurdles that can, and should, be overcome.
To learn how you can take advantage of new technology and strategies, and take OT to the next level, watch our Industrial IoT webinar "Industry 4.0: Predictive Analytics and Its Immediate Impact Across Industries." It features IoT thought leader, Stanford Lecturer and Alchemist Accelerator Chairman, Timothy Chou.
----------------------------------------------------
Thanks!
Erik Martinez