Immelt heralds marriage of materials science and data analytics

19 Jun 2013 | News
There is an "massive economic opportunity" to use a combination of materials science and data analytics to improve fuel performance and reduce downtime in industry, according to Jeffrey Immelt, chairman and CEO of industrial conglomerate GE

The analysis of data collected by scores of sensors combined with materials science will enable companies to predict when a machine or component will fail and take proactive action, Immelt told delegates at a GE event on the "Industrial Internet" held in London's derelict Battersea Power Station (see related new story for more on the Industrial Internet concept).

GE is looking to use data analytics to eliminate downtime and optimise the performance of both assets and enterprises, Immelt said. "Small performance changes can drive massive economics in the marketplace and for our customers," he claimed.

He described the Industrial Internet, along with the unconventional energy revolution, advances in manufacturing and the opportunities IT offers for internal improvements, as one of four major economic opportunities for GE.

Immelt estimated that the ability to make better decisions faster will generate 90% of the productivity gains arising from the Industrial Internet.  In the healthcare sector, for example, Immelt predicted that instant access to best practice and the latest protocols will be pervasive within five years, meaning physicians everywhere will have access to completely up-to-date information.

"Data is a new economic resource, a new raw material," added Ken Cukier, data editor of The Economist and speaker at the GE event. Cukier outlined how computers can detect patterns in large volumes of data, enabling them to predict when a machine will fail, detect the first signs of cancer or when a driver is losing concentration and may be about to crash. A shift in posture could signal driver fatigue, prompting the car to vibrate the steering wheel and warn the driver to watch the road, Cukier noted. He said that, in theory, the incumbents in any particular sector have an advantage because they have the data.

However, there are also risks associated with big data, particularly around the protection of privacy and propensity. Cukier pointed out that algorithms may be used to predict what we will do and then penalise us for something we haven't yet done. He argued that we need to combine our humanity and common sense with big data and analytics to ensure that these tools are used in appropriate ways.

In response to a question from Cukier, Immelt described the Industrial Internet "as one of [GE's] big investable themes" comparable to its early focus on emerging markets and investment in oil and gas technologies over the past decade. Immelt also noted that the continuous flow of data on the condition and performance of GE's products means companies can no longer have arm's length relationships with customers: Instead, the relationship will be performance-based. "We will need new business models, we will have to experiment with risk-sharing," Immelt added.

Never miss an update from Science|Business:   Newsletter sign-up