It is easier than ever to store massive amounts of data. The cost of wireless sensing technologies has dropped dramatically, making it much easier to instrument every asset in the plant and ensure that no performance aberration goes unnoticed. However, bringing in all this data carries with it the risk of information floods. If data is collected and stored without a plan, it can easily become overwhelming.
To help organizations better manage their data, Emerson’s Brian Overton and Drew Mackley recently shared four condition monitoring best practices in an article in Processing magazine. These tips help maintenance teams quickly and easily turn condition monitoring data into actionable information to help plant personnel make better decisions.
Store what you need
Today there is practically no limit to how much data a maintenance team can store. Teams that were once starved for data now find themselves swimming in it, and it is often tempting to collect and save everything. However, saving every piece of condition monitoring data presents problems. Not only do storage costs quickly escalate, but it also becomes much harder to find critical data among all the insignificant data points.
To avoid information overload, teams can develop a storage strategy based on asset criticality. When teams use a criticality strategy, the most critical assets (typically those for safety and those necessary for continuous operation) collect and store data more frequently than balance of plant equipment (Figure 1).
Figure 1: Collected data is easier to manage when based on asset criticality.
Predicates eliminate unnecessary data
Many assets in the plant run intermittently and collecting and storing data on those assets when they are not running can build large collections of unnecessary data. Teams using predicate condition monitoring only collect data from assets under specific machine states. Not only does this help them eliminate data floods for dormant equipment, it also empowers the team to collect data under specific states to more easily identify the root cause of issues that only occur under certain circumstances.
Drive better results with edge analytics
Few plants have the time and resources to collect large amounts of data, store it, and wait for manually analysis. Today’s best continuous monitoring technologies offer edge analytics to eliminate much of the manual labor behind turning data into actionable information.
In today’s environment of limited and inexperienced personnel, analytics at the edge can ensure critical issues with plant assets don’t go unnoticed.
Figure 2: Predicate rules help eliminate low- or no-value data.
Bring it all together
If data is siloed in many different systems across many different areas, it is difficult to draw useful conclusions about the holistic health of the plant. When teams bring data together into a single, persona-based application, they can more easily view data, understand it, and collaborate to improve the health of the plant.
The most advanced applications not only offer comprehensive visibility into plant health from a single dashboard, but also provide mobile tools, to help plant personnel maintain awareness from anywhere they may find themselves.
Emerson drives more efficient condition monitoring
Emerson has a wide range of continuous condition monitoring tools to help maintenance teams make the most of their limited time and personnel. One example is Emerson’s AMS Asset Monitor, which can be quickly and easily installed right at the site of an asset and performs analysis on vibration data with results that are viewable on any device using a web browser. Even if a site has no analysts, the intuitive data makes it easy to instantly understand the health of any asset in the plant.
Other tools such as Emerson’s AMS Machine Works bring all monitoring data into a single location with a standard interface to break down silos and drive more value from condition monitoring data.
To learn more strategies for improving your data collection, and to see some specific examples, read the article in its entirety at Processing magazine. And while you’re here, I’d love to hear your strategies for eliminating data floods and making the reliability information you collect as useful as possible. Please feel free to comment below with your ideas.
