General Electric to develop large data application plan

With the aid of sensitive data collected from industrial products manufactured by GE, a new generation of analytics technology is about to create more opportunities for a range of industries.

William Ruh, vice president of General Electric software and analysis, elaborated three key components of the company's big data strategy in December 2012: Predictive analytics, cloud platforms for analysis, and security.

As one of the world's largest companies, GE has been a major system producer in aviation, railways, mining, energy, healthcare and other industries. After recognizing the importance of big data for the company, GE CEO Jeff Immelt launched an innovative project called "Industrial Internet" designed to help customers improve efficiency while using analytics to create new profit opportunities for GE.

The "Industrial Internet" is an important step in GE's foray into the internet of things, and it can collect large amounts of data and analyze it through the sensors connected to it. Immelt says they have embedded sensors in 250,000 "smart machines" manufactured by GE, including jet engines, power turbines, medical devices, etc. Acquiring and analyzing the data generated by the sensors has great potential for optimizing industrial operations.

For example, GE intends to analyze the alarm and error reporting data they have collected over the past five years and use the data to identify patterns of behavior that can make planning maintenance more accurate.

In the aerospace sector, GE's GEnx jet engine analyzes 5,000 data points per second to optimize flight time. In the area of rail traffic, GE and customers can determine which wheels need to be replaced by collecting and analyzing data from sensors. At the same time, they plan to add some cameras to record video about the tracks and analyze the video in real time.

GE also plans to use data collected from other sources to enrich their datasets. For example, use satellite data to enrich smart grid maps. In this way, the power companies know where to remove the vegetation can reduce the risk of wires in the storm short circuit.

Ruh said GE will use existing vendors and open source components to achieve breakthroughs. While the company will use technologies such as Hadoop and R, he also made it clear that "no single vendor has a panacea for the problem". He says that machine learning is a unique area and he wants his team to make some new breakthroughs in this area.

This is done, Ruh wants his technical team to be able to create an analysis cloud so that the relevant tools can be accessed at any time for data collection and analysis. The platform also requires sufficient flexibility to enable enterprises to migrate to emerging technologies in the future. Finally, security and compatibility are important parts of GE's attention. One of the most important things is the ability to monitor who has done what at what time. This move extends the forensics technology in computer systems to industrial systems.

Scalability, Ruh says, is a fundamental architectural challenge across the entire field. According to him, traditional enterprise software for business transactions does not apply to real-time, high-speed, high-volume data generated by the General Electric system. For example, a key requirement is to store time series data that incorporates a similar Hadoop analysis technique.

To achieve this, GE's more than 9,000 employees around the world are developing the software. However, research and development of software and analysis technology is only part of the company's basic work, this work in Immelt seems to need to further deepen. Immelt has said GE wants to set up an office in the San Francisco Bay Area to make it easier to recruit engineers with analytical expertise and work with start-ups. In addition, GE will take a similar approach to investing in health care companies to acquire stakes in some companies.

In addition, GE established a software and analysis headquarters two years ago in Santa Lamon, California, near Silicon Valley, and successfully dug Ruh from Cisco to lead and launch technological innovation. Ruh says the office has added 300 employees, more than half of them developers, and the rest are architecture, marketing, support and operations staff. Ruh hopes that by the end of 2013 the team will be able to increase the number of people to 500. (Gundy compilation)

(Responsible editor: The good of the Legacy)