By Achmad Chadran, Portfolio Manager
This is part two of a four-part series!
There’s no single definition of success
Do you define Predictive Analytics success in terms of increasing operating efficiency, or its sidekick, decreasing operating costs? Or does success to your organization mean the ability to make better use of fewer plant outages? Maybe capital expenditures are more your thing, and you want an Asset Performance Management (APM) solution that helps you improve ROI from your infrastructure.
My point is this: Predictive Analytics can be a powerful tool, but unless you have a clear idea of what you hope to get out of it, you may set yourself (and your organization) up for disappointment. Note too that I’m not saying that you can’t have all of these as success criteria for your solution.
Whether it’s all of the above, some combination thereof, or a definition not even on this shortlist, your definition of success needs to apply uniquely to your operation.
So how do you go about defining success?
Formulate a consensus view
First of all, let’s be clear: success probably won’t be yours alone to define.
Still, finding and implementing a Predictive Analytics solution is an initiative you’ve taken on, so go ahead and document your personal ideas about your priorities and goals.
Next, identify the people in your ecosystem who’ll have a stake in solution success.
There may even be people outside of your organization who can help you define success. Think: partners, regulators, and even strategic (and trusted) vendors.
Now, as you reach out to your broader community as a sounding board and brain trust, you’ll also want to build a smaller cross-functional team to work with during the shortlisting, vetting, piloting, and selection processes.
Be sure you build a cross-functional team that is motivated to help you conduct research, set those priorities and goals, and then apply these objectively to the vendors who’ll want your business.
Set concrete goals
At this point, you’ve got the vision and you’re building the consensus. With hope, you’ve already started defining and assigning tasks. Pat yourself on the back: you’re on your way!
My next bit of advice may seem obvious, but it’s absolutely vital: make sure you set project goals that are both timebound and measurable. Don’t settle for abstract goals like “improve productivity” or “reduce unplanned outages.” Instead, formulate goals more along the lines of “reduce unplanned outages by 25% within two years” or “achieve $1 million in savings during Year One from extended production times, reduced overtime pay, reduced repair/replacement costs, and extended asset lives.”
Clearly, goals like these require you to benchmark current performance.
Your goals may also include more corporate governance aspirations, such as a 10% reduction in safety incidents or targeted reductions in regulated emissions Predictive analytics can help you achieve any or all of these, provided you define these goals from the outset. In a way, the specific goals you and your teammates choose will be less important than the metrics and timeframes you apply to them.
Follow these high-level guidelines to ensure predictive maintenance success.
It’s great to have a vision for how Predictive Analytics can help improve the efficiency and effectiveness of your operations, but it’s critical to get a consensus view on what success looks like. It’s also critical to get buy-in from peers across different work teams and organizations, and to have people who can assume task responsibility and help maximize transparency and accountability.
Your work to define requirements, priorities, and concrete goals could well determine whether you implement a solution that materially improves operations or becomes nothing but a costly debacle.
Coming up: Embracing the Search
Join me for the next installment in this series, where I’ll offer advice on conducting the actual solution search. Go ahead and subscribe to our monthly blog updates.