Ge Predictive Case Study

Unlock access to all the studying documents.

View Full Document

Applications in Predictive Analytics: A Case Study on GE

Shaun Lara

Northwestern University

This paper explores the evolving technology and fast-paced method a major company is

using to solve customer problems, stay relevant with their market, and make money. General

Electric touches many industries and therefore has a wide range of applications for using

predictive analytics in asset and operations optimization. We will take an early look at the value

GE’s adoption of analytics is bringing to it’s business. Also, how GE is acquiring and

effectively using its data for its wide-ranging market will shed some light on their business

practices.

Behind all this technology that General Electric is introducing into their offerings is lots

and lots of data. Many of the markets and businesses that GE provides products and services for

are already giving off uncounted amounts of data. GE’s vision is that everything it makes from

plane engines to wind turbines to MRI scanning machines will have sensors and RFIDs that will

be able to capture this currently unused and undiscovered data. For example, “a single gas

turbine sensor, for example, creates 500 gigabytes of data daily. With approximately 40,000 gas

turbines operating worldwide—and assuming three sensors per gas turbine—60 quadrillion bytes

of data would be generated per day (24 times the daily traffic generated by the global Internet in

2000). That’s just for one sector.” (The Industrial Internet, 2013) GE is also using exploratory

data analysis to discover what the data can tell them beyond their hypothesis testing. This has

led to new unrealized Predictivity efforts to enhance productivity that they, or their customers for

that matter, didn’t even know existed. GE has to make sure the data they get is measured

accurately with advanced sensors. Improvements in technology today and the power of

economies of scale have allowed these advanced sensors to cost less than $10 a piece. (The

Industrial Internet, 2013) In addition to these advanced sensors, GE claims to also maintain

extensive data error logs to identify and eliminate the root cause for errors. In fact, there is a

whole division within GE Predicivity dedicated to these efforts.

According to a recent posting from wikibon, the Industrial Internet is expected to provide

a value to business of $1.2 trillion, up from $23 billion last year. (Defining, 2013) This is the

ultimate value GE is hoping to capitalize on with its Predictivity and analytics efforts. This kind

of transformative change requires fundamental changes to a business’s management structure in

order to swiftly operationalize the strategic insights collected from analyzing all this newly

discovered data. As mentioned earlier, with buy-in from the highest possible levels at GE

(Immelt), they have created a new executive structure around their analytics endeavors and

imbedded them within each of their industrial areas. GE also realizes that in order to execute an

analytics business successfully they will need outside help from subject matter experts who can

fill in the gaps of the capabilities they require. Particularly in the areas of software development

and cloud-based capabilities. GE has partnered with Amazon Web Services, Pivotal, and

Accenture in order to meet their big data and analytics needs. One would surmise that with a

company the size of GE they would eventually learn to do these operations in house and more

cheaply, but the need to get to market quickly requires this approach. In regards to GE’s alliance

with Amazon, Werner Vogels, Amazon.com CTO, offered a short statement outlining the

motivations behind the new alliance: “Decades of GE-led innovation have helped shape history,

and we are excited to partner with the GE team to help shape the future of Industrial Big Data by

helping GE bring together intelligent machines, advanced analytics and industrial applications

using the AWS cloud. GE’s domain knowledge and R&D capabilities combined with the

strength of our global infrastructure, operational excellence and breadth of services will enable

customers to solve problems in ways we haven’t even imagined yet, such as improved accuracy

in healthcare treatments or extreme levels of energy efficiency.” (Wheatley, 2013)

So far in our readings, we have examined what Predictive Analytics is to an industry and

the power this technology can bring to improve a company’s goods and/or services. In Eric

Siegel’s Predictive Analytics: The Power to Predict Who Will Click, Buy, Lie, or Die, he

provides a thoughtful definition in his introduction on what Predictive Analytics is: “Predictive

Analytics is the process by which an organization learns from the experience it has collectively

gained across its team members and computer systems. In fact, an organization that doesn’t

leverage its data in this way is like a person with a photographic memory who never bothers to

think.” One company with a lot of fingerprints on different machines, processes, and business