3 Tests for Data Criticality in EAM: What Stays and What Goes
Use these three tests to determine if the data you have is actually the data you need.
There is such a thing as too much data, even when developing data-rich proactive asset management and maintenance strategies.
In a perfectly balanced asset environment, the best actionable information will tell operators how equipment performs, when and how equipment underperforms, the risks to production should an asset fail and the recommended resources for resolving the issue. But sometimes the data that asset-intensive businesses collect is actually worth far less than what's expended acquiring it, conditioning it and storing it. According to a study from The Economist, 6 in 10 manufacturers are unsure they can keep pace with their current data collection processes. In fact, half are unsure they can draw any valuable conclusions from the data they have, as "it comes from too many sources and in a variety of formats and speeds."
As important as complex data sets are
1. Does the data align with my organizational goals?
Often asset-intensive businesses fail to do anything meaningful with the data they collect because they first failed to set quantifiable high-level objectives.
"Reduce downtime" is so vague, it's hardly a goal at all. "Reduce downtime by 15 percent" is a little better. "Reduce downtime on our three most underperforming lines by 15 percent by the end of next year" is the best of the three. As business leaders elaborate their broader goals and tease out more specific details, the data required to confirm them will practically present itself.
Apart from data synced directly with greater organizational strategies and improvement efforts, other vital data has one or more of the following earmarks.
2. Does the data pertain to scheduling preventive measures?
Proactive EAM is a balanced combination of the preventive and the predictive (with a little prescriptive thrown in for good measure, thanks to cognitive EAM solutions).
Valuable preventive maintenance data includes any information that would trigger an alert for maintenance unrelated to real-time performance: an expired bearing, a belt that has exceeded its nth rotation, a subassembly that has passed its warranty, etc. Although no mechanical failure has occurred yet, each of these events would lead to at least an inspection and at most a work order. Less obviously, this also includes resources such as historical data on equipment, operational
Preventive measures, however, are not created equal. Since PM can consume much of a maintenance professional's workday, companies must leverage asset criticality rankings to determine which PM-related tasks take precedence over others.
3. Does the data describe mechanical condition?
Condition monitoring is a linchpin of real-time predictive maintenance, so the importance of gathering real-time information regarding equipment excellence cannot be overstated.
But manufacturers must be cutthroat about the sorts of operational data they pay attention to. Advancements in automated industrial equipment and robotics have exploded the possibilities of what asset-intensive businesses can learn from their equipment. Without a substantial, goal-aligned plan for utilizing environmental data, supported by failure modes and historical performance, companies can easily pack away gigabytes, even terabytes, of data daily. And there it will sit in its servers, soaking up electricity, wasting data center spend and obfuscating truly actionable data.
Always ask: What does this data tell me, and how does that prevent failures and deficiencies on my most critical assets?
For more information on data management best practices, contact the enterprise asset management professionals at ABS Group.