Hottest sensor big data Ge builds industrial Inter

Sensors + big data: Ge builds industrial interconnection

GE power turbine that can send data on power consumption

with big data and sensors, Ge hopes to build an interconnection of industrial systems to further improve energy efficiency

ge (General Electric) has a loud name for future business industrial interconnection. This is a term coined by GE's R & D department. Ge hopes to obtain massive data by adding more sensors to its products, and ultimately help the company improve the energy efficiency of its locomotive and aircraft engines, nuclear magnetic resonance instruments and other equipment

ge plans to invest $1.5 billion in industrial interconnection projects in the next three years, part of which will be used to support the research projects of the newly established software R & D center in San Ramon, California. For example, Anil Varma, a machine learning expert at the research center, is experimenting with how to screen the imperceptible alarm signals in 20000 jet engines produced by Ge to predict which equipment needs maintenance. For some types of engines, Varma's algorithm can predict their maintenance needs one month in advance, with a prediction accuracy of 70%, which can greatly reduce flight delays. (editor's note: the crew has been notified of engine failure more than once before boarding)

Franz Spitznagel, senior development project manager of DuPont performance materials (DPM), said: "In the power field, big data engine

in the past, the sensors in GE's aircraft engine were in passive mode, and the red light on the instrument panel would not be on until there was a fault. There are many such sensors, such as measuring temperature, pressure and voltage, which were rarely retained and studied in the past. In most flights, the engine will only retain three average values, namely takeoff, cruise and landing data.

according to Varma's introduction , GE's next generation genx engine (equipped with Boeing 787 aircraft) will retain all basic data of each flight, and even transmit it back to ge for analysis in real time from the aircraft. The amount of data produced by such an engine in a year will even exceed all data in the history of Ge aviation business

although communication between machines through sensors is no longer a new concept (such as IOT), GE's business scale can make this idea faster. We have the largest industrial data set because we have been operating these devices for a long time, Varma said. We also have historical data and monitor future data, which allows us to test the feasibility of any algorithm

according to William RUH, a former Cisco executive and head of GE's industrial interconnection project, GE's industrial interconnection concept was born several years ago during the financial crisis if there were special requirements. As the uncertainty of economic growth increases, industrial customers begin to turn their attention from improving productivity to improving profit margins. The concept of big data is also becoming more and more popular. Finally, William RUH's team began to consider whether to formulate a data strategy for GE products

1% beyond the physical limit

Venkatasubramanian, a professor at the school of data science and engineering at Columbia University, believes that GE's application of big data technology to solve industrial problems may not be smooth sailing. For a commercial retail company, it is enough to find the correlation between consumer data, such as the famous beer diaper theory. In this primary application, the current standard machine learning algorithm is competent. But for complex physical systems, the data model also needs to be able to explain the reasons behind the correlation

researchers from GE software R & D center demonstrate the data visualization tool of the power plant

at GE's San Ramon software R & D center, researchers are developing new user communities to become stable growth, structural adjustment, mode transformation, and efficiency benefits, and help people visualize Industrial data through maps, simulation, and twitter like device socialization. One of the laboratories has many large screen displays connected to Microsoft's Kinect somatosensory game equipment. Workers in the power plant can interact with the data visualization interface through gestures to help make operational decisions about regional electricity

ge also worked with a Canadian power company to predict the hot spot areas of tree pruning by analyzing satellite images, weather maps, local power failure records and other data (falling branches are one of the main reasons for power failure caused by lightning)

in addition, Ge also cooperated with a medical center in New York City to implant sensors in hospital beds and medical equipment to reduce the empty bed rate and improve patient reception capacity

our operating efficiency can be improved by 1% when household appliances and automotive related products can't be painted or painted in time, but this goal can't be achieved through better equipment, because we have brought physics to the extreme. RUH said

obviously, GE has the answer, that is, the industrial interconnection of sensors + big data