See what's coming next—and get in front of the curve
Big data and predictive analytics are usually associated with strategic marketplace decision-making, such as determining which products to offer which customers at which times.
But did you know these tools can also drive asset infrastructures to new heights of performance and uptime?
That's a truly compelling prospect for most organizations, which are increasingly dependent on the infrastructure to deliver more and more services—both internal and external. Every year, the costs of unscheduled outages and maintenance continue to climb. And every year, organizations try to respond more quickly and effectively to keep the costs low and the outages few.
But really putting a dent in operational downtime and performance shortfalls means accomplishing the seemingly impossible: accurately predicting in advance what's going to break, when, and what the expected impact will be.
Need a real-world example? Consider the case of energy providers, which are typically reliant on mechanical turbines. When the turbines fail, electrical output declines, and for everyday citizens that's a situation that can literally mean the difference between life and death in sweltering summers or gelid winters. What if, instead of reacting to a turbine problem more quickly, energy providers could anticipate it before it happens and proactively avoid it altogether? The costs and total impact would both fall to zero—and the time and energy spent solving it could instead be dedicated to creating entirely new forms of value.
Big data and predictive analytics can create the crystal ball organizations need to make that happen. By bringing together sensor-sourced information about asset status and performance, and business intelligence to assess the changes and prioritize the response, the seemingly impossible becomes quite possible after all. Future problems can be spotted, like obstacles in the road ahead, and once they're spotted, it's a much simpler matter to steer the organization around them completely—or guide the organization through an obstacle course in an optimized way, if it turns out there are many obstacles.
IBM Predictive Maintenance and Quality: Apply the power of big data analytics to the asset infrastructure
The included Decision Management methodology empowers you to find an efficient balance between resource constraints and costs. This methodology can, by incorporating organization-specific information like asset and process rules, help you accomplish your business goals without breaking the bank.
It was with just this premise in mind that IBM recently introduced a new solution: IBM Predictive Maintenance and Quality. This offering—powered by advanced analytics drawn from IBM's SPSS and Cognos solutions and integrated with the IBM Maximo asset management family—delivers three key capabilities:
Let's walk through some of the specifics of the solution to understand how it accomplishes these goals.
Real-time monitoring and reporting
Before you can accurately predict how assets will perform, it's first necessary to determine current asset status and performance. If you can't see what's happening to individual assets, you can't extrapolate how well or badly they are performing against goals, and you can't determine which business processes or services are likely to be affected if the asset fails.
That's why IBM Predictive Maintenance and Quality can draw information from a remarkably wide variety of assets, in a wide variety of ways. These include via sensors, power line communication, SCADA (supervisory control and data acquisition) systems, databases, maintenance logs, and Big Data streaming sources. Once the various information streams are aggregated, they can be analyzed, looking for patterns or trends that suggest a forthcoming issue.
Excessive heat buildup in a particular area of a data center, for instance, might suggest higher odds for a service outage, due to a failure of physical hosts in a too-hot environment. Or SCADA data, properly interpreted, might show that water is moving from point A to point B in a municipal network slower than it should, and more slowly over time—implying a complete blockage will occur unless steps are taken.
Big data and advanced predictive analytics
Of course, monitoring per se is only the beginning of the story. Most complex trends involve many assets, and thus are difficult to spot; disparate data sources have to be correlated in a shared context before the pattern emerges.
Hence the inclusion of Big Data/predictive analytics capabilities in this offering as well. They provide just such cross-domain analysis to deliver three kinds of insight: descriptive (what's happening?), predictive (what's going to happen next?), and prescriptive (what should be done?). They also provide data and text mining functions.
Asset health is both evaluated and predicted based on data such as logs, alarms, and metrics/measurements, also taking into account the asset's history of repair requirements. In this way, organizations can accurately understand not just which assets are up and running, but also which will likely continue to be, and which represent high-probability near-term failures, and demand a proactive fix.
If that sounds complex to manage, it's not. No custom code needs to be written in order to achieve these goals. The menu-driven GUI is all that's required to create predictive models that will accurately reflect the real infrastructure and anticipate its biggest challenges in the foreseeable future.
Quick and accurate decision-making
Once you understand what's likely to happen, the next question is: What should you do about it?
Answering that question is not as straightforward as it might seem. While it's common to talk about "the optimization of a complex asset infrastructure for best business value," actually fulfilling that goal, by making the right decisions at the right times in the right ways, is no simple feat.
Fortunately, IBM Predictive Maintenance and Quality includes a number of features to make it as simple as it can reasonably be. For instance, the included Decision Management methodology empowers you to find an efficient balance between resource constraints and costs. This methodology can, by incorporating organization-specific information like asset and process rules, help you accomplish your business goals without breaking the bank.
The solution also incorporates "what if" simulations, so that managers can consider many possible scenarios, assess their outcomes, and choose from them the best available candidate—one that will meet needs even if conditions change unpredictably.
What's the fastest way to get up to speed on maintenance status in the infrastructure—or perhaps across multiple infrastructures, at multiple sites?
Answer: the predictive maintenance scorecards provided by IBM Predictive Maintenance and Quality. These give at-a-glance, color-coded insight into status and health, as drawn from virtually any data source. Self-service queries give you the power to dial in the results you want, drilling down from higher levels of abstraction to lower, whenever that's needed.
For even more customization, check out the studio environment; it uses a drag-and-drop interface to give managers an exceptionally configurable perspective. And for similarly impressive convenience, the same information is available from smart mobile devices—so managers can attend to basic monitoring tasks anywhere they go, any time they want, and on almost any platform.
Seamless integration with IBM Maximo asset management
Another strength of IBM Predictive Maintenance and Quality: the fact that it supports direct integration with certain enterprise asset management systems including, of course, IBM Maximo.
What does this mean? To begin with, Maximo becomes another exceptionally rich data repository for IBM Predictive Maintenance and Quality to draw upon. Specifically, master data is synchronized between the two solutions, so the smart predictive analytics discussed earlier can be extended in new directions directly applicable to asset maintenance and optimization.
And should IBM Predictive Maintenance and Quality find, based on its analysis and business rules, that some sort of action should be taken, it can actually initiate that action—spawning new work orders appropriately.
Essentially, the two solutions complement each other perfectly—each benefiting from the other's particular capabilities to leave the organization better prepared to get the highest possible value from all assets, in all domains, both in the present and the immediate future.
Accelerated time to value
In an increasingly fast-paced business arena such as today's, time is money—and the less time required before a solution can start contributing in a practical way, the better.
IBM Predictive Maintenance and Quality delivers exceptionally quick results via a number of design strengths. First, its business user interface supports fast master data entry and modification; this is critical information needed for practically everything else the solution does. The installation process has also been engineered to be particularly fast and straightforward, thanks to preconfiguration in certain cases that speeds integration with the rest of the infrastructure.
Many organizations will find that right out of the box, the solution's data source connectors and models, dashboard, and reports—designed to match common business requirements and contexts—will serve, with little to no customization needed. And in the event the included content does need to be modified or extended to support special industries or business applications, that too is a swift process.
The more data sources that are linked to IBM Predictive Maintenance and Quality, the more trends and patterns it can detect, and the better the overall visibility into maintenance issues it will deliver. That's why IBM has gone out of its way to ensure this solution has an unusually open architecture.
For instance, as noted, enterprise asset management solutions integrate with it. Data can be streamed directly from many sources. And there are also integration APIs (application programming interfaces), to let those who are willing to do a little coding increase the list of data sources even further.
Suppose, for instance, it turns out that IBM Predictive Maintenance and Quality's capabilities might be needed to analyze data that is currently locked up in a homegrown application. The integration APIs can be used to create a logical bridge by which that data can be made available, allowing IBM Predictive Maintenance and Quality to work its magic in an entirely new way not even IBM has imagined.
Click here for a demo of IBM Predictive Maintenance and Quality.
Harness the value of big data to optimize smarter infrastructures
Big data analytics, coupled with the right facilities and asset management software, can provide next-generation opportunities to improve facilities and asset management processes and returns - increasing revenue, powering operational efficiency, ensuring service availability and mitigating risk.
The next generation of application monitoring has arrived! IBM SmartCloud Monitoring - Application Insight is purpose-built for public and hybrid clouds, provides visibility into what matters the most – the end user experience. It is lightweight, scalable, and able to be deployed in minutes without installing anything in the infrastructure.
Best-in-class Patch and Power Management solutions for only $5
Significantly lower the total lifecycle costs of managing and securing your computers for only $5 per endpoint.