Cooling Down the Summer With Predictive Maintenance
Brittany Detamore 270006ARTU email@example.com | 2013-07-11 13:22:58.0 | Tags:  predicitive-maintenance analytics energy outage brownouts enterprise-asset-manageme... big-data predictive-analytics power | 6 Comments | 10,152 Visits
Follow Anuj on Twitter @anujmarfatia
As with any excessive heat, my air conditioner was working in overdrive and our freezer was opened several times as I basked in the cool, refreshing air. As the dials on my electricity meter were spinning out of control and I could already imagine the upcoming electric bill, I was thinking that others were probably in a similar situation.
What worried me was not the fact that I couldn’t get the lyrics, “It’s a cruel, cruel, summer,” out of my head (though that is annoying), but that if everyone was blasting their air conditioners, was the energy grid infrastructure reliable enough to not cause brownouts, rolling blackouts or an extended power outage?
There have been too many such occurrences – everything from New York City’s 2009 heat wave that caused 3.2 million customers to lose power to a heat wave in India in May 2013 that led to grid failure and caused over 500 deaths to most recently power outages caused by heat in San Jose right before the July 4th weekend where over 9,500 customers were affected.
According to the U.S. Energy Information Administration, "enough electricity must always be produced to meet demand at every moment." When power grids are unable to produce enough power (because of overwhelming demand), blackouts can happen.
Most electric generation is driven by turbines –rotary mechanical devices that extract energy from gas, steam, or water flow. When these turbines fail completely or do not perform as they were intended, enough power cannot be transmitted to consumers. Many of the turbine failures are unplanned. While many power generation companies utilize weather forecasting techniques in some manner, trying to ensure they optimize usage of their limited assets or turbines is another story.
Most organizations respond to turbine failure by sending out maintenance crews, though the process is very reactive. And organizations in other industries deal with similar problems with their assets- including a manufacturing line, an offshore oil drill, a city water pipe or a mining excavator - and respond in a similar manner as well.
In order to ensure that such “surprises” are limited as much as possible, IBM has created a cross-industry, packaged software solution – IBM Predictive Maintenance and Quality – that can predict when and how an asset is going to fail.
The software solution includes data integration and analytics, as well as decision management technologies, so organizations can capture all of their Big Data, garner insights and predictive outcomes from that information, and quickly and accurately make the right decisions at the right time.
The solution not only monitors asset sensors – it has the capability to leverage input from other sources such as environmental, facilities monitoring systems, and even text information such as maintenance logs. The solution also tightly integrates enterprise asset management (EAM) and analytic systems capabilities to offer distinctive benefits across an organization’s enterprise. The packaged solution also includes data connectors, data schema, predictive models, dashboards and reports - accelerating the time-to-value for organizations.
Organizations using IBM Predictive Maintenance and Quality will be able to:
For instance, IBM and a business partner, Genius Systems, worked with Israel Electric Corporation (IEC), the primary energy provider in Israel. IEC was able to model the behavior of its turbines, and monitor their performance in real time. When anomalies are detected, it can quickly trigger maintenance resources to fix problems before outages occur or efficiencies are reduced. Specifically, IEC was able to reduce costs by up to 20% by avoiding the need to restart turbines after an outage and utilize an early warning of certain types of failure up to 30 hours before they occur, instead of 30 minutes.
To hear more about IEC, as well as the topic of predicting asset failure, please listen to this cool webinar, which includes a panel from Forrester Research, Genius Systems, Israel Electric Corporation, and IBM.
For more information: