The following is a contributed article by Bryan Friehauf, EVP and GM of Enterprise Software Solutions at ABB.
When big data took off and utilities began taking advantage of all the information it had to offer, many believed this was the end of guess-based decision making and the elimination of risks. After all, utilities finally had all this great information at their fingertips! But while rapid technological innovation has given the industry more access to data than ever before, it's not just about having the most data.
The key to success is having the right systems and processes in place to infer the right insights from that data. Data processes are as important as the data itself.
For utilities, asset performance management (APM) puts in place these processes on the data and acts as a new age "fortune teller." APM usually focuses on data collection and processing. From there, operators must collaborate on the data to add context. This, in turn, improves asset management and takes the guesswork out of the equation.
Today's fortune telling capabilities
In the last several years, equipment operators have made great strides in early detection of equipment degradation prior to the function failing. Adding sensors, communications equipment and analytics software to assets, as well as the ability to tap into historical data (for example, inspection and maintenance data a company has collected over many years) has accelerated this predictive forecasting.
Now, operators can collect more data more frequently, automatically, making asset monitoring more effective. Adding a layer of analytics to the data has turned the predictive models into prescriptive models that recommend corrective and preventative actions to take on a degrading asset.
Using this model, operators can monitor and manage more than just one asset at a time. As utilities are afforded more time to make decisions on asset health, and they seek to optimize risk against life cycle cost, aggregating all asset data becomes necessary to maintain the total system health versus just one individual asset. This aggregation can be complex, with multiple system configurations and dependencies.
A full system perspective also requires more than just direct condition data, but also a review of all indirect business process trends.
To create a reliable system, operators must manage both the asset's physical conditions and the business process. Business process trends like compliance, backlog size, obsolescence and spare parts availability, operator workarounds, near misses, and operating experience can complement direct information on current equipment condition and historic performance for a comprehensive, aggregate view of health on a medium to long-term horizon.
In the era of big data, aggregating direct and indirect data at a functional system level and the development of collaboration and action tracking processes are proven process solutions that must be incorporated into an optimized APM strategy.
Tomorrow's fortune telling capabilities
Over the next ten years, organizations will incorporate new technologies and communications into their work, further evolving the predictive maintenance paradigm. As a result, we anticipate that automated algorithms will be able to take actions independently and cover key equipment and processes.
As it stands today, algorithms are not comprehensive enough to tackle this and are only capable of managing specific tasks. This means organizations are still unwilling to fully trust them.
For organizations to get fully on board, they need to be able to make changes to the automated predictive analytics framework. This will help users understand the context and create a complete picture for decision making. It also helps them account for and reduce risk.
Human interaction with the data will help algorithms improve and become more comprehensive and independent in their analysis, recommendations and automated actions.
Getting the team on board
No matter how sophisticated the predictive analytics and system health perspective, operators tend to identify more risks and required actions than they can reasonably resolve or even mitigate given their allocated resources. Often, the insights from data are overwhelming, and operators are unable to make decisions as a result. The systems and the processes we use can help manage risks with time and costs.
Typically, many silos across the organization are involved in the workflow of risk identification and task prioritization. But this workflow can offer some of the greatest opportunities for streamlined cross-silo collaboration. To realize the value from APM processes, connectivity among APM, work, and operations management tools is key.
Additionally, organizations should provide role-based interfaces for these different silos to share information and priorities. When all the various groups have a stake in the full outcome, it's easier to get the buy-in from the whole organization that allows for the successful completion of tasks. Collaboration produces better decisions and the alignment and buy-in that comes with it drive effective execution.
Not all decisions based on predictive probabilities will tell our future. Over time, however, consistent application of such decision-making processes in APM pays off in higher reliability and lower costs.
And, with a collaborative process supported by human interaction with the data, the system encourages effective risk-based decision making, builds in safeguards, discourages procrastination on maintenance and other preventative work, and secures organizational buy-in for technological transformation.