Industrial big data: Core competiveness of manufacturing enterprises in the future

At the Big Data and Intelligent Manufacturing Salon held in Changsha, China, experts from various disciplines discussed how industrial big data will shape future manufacturing and bring new services that focus on customization and commercial needs.

By Stone Shi June 6, 2016

The Industrial Big Data Application and Intelligent Manufacturing Salon was held in the Sany Heavy Industry headquarters in Changsha this month. The salon gave participants deep understanding on how to create intelligent plants in the Industrie 4.0 era by using big data.

One of the first intelligent manufacturing demonstration projects in China is the Sany No. 18 factory building. The total area of Sany No. 18 factory building is about 100, 000 square meters, and there are multiple assembly lines: concrete machinery, pavement construction machinery, and harbor machinery, etc. Digitalized factory emulation technique has been applied for scheme design and verification. In this factory, the traditional discrete manufacturing model is abandoned, substituted by high-flexibility, mixed loading of multiple product types and flow mode.

"This is a smart factory building integrating [a] large-scale computing system, traditional operating tools, and large-scale production equipment. Each production process, quality test, and the workload of each worker will be recorded; over 1TB data will be produced weekly for the factory building; the data is used to boost production and improve intelligence and flexibility. Multi-type and small-batch production can, in this way, be achieved. This will improve efficiency, reduce cost, and guarantee quality," said Dongdong He, the vice president and chief process information officer of Sany Heavy Industry.

"We perform networking and status monitoring for 200,000 engineering machineries; 200 million pieces of data are uploaded each day, and now the accumulated big data resource has exceeded 40TB. These data can help us learn the equipment position and track real-time working conditions to monitor early warnings, conduct remote fault diagnosis, and improve predictive maintenance.

It [can] help us improve the product quality to win the market," said Dongdong He. "We can also provide finance services based on customer and equipment usage data, expanding the innovation to the commercial side. Therefore, I can firmly say that industrial big data will become the core advantage of Sany Heavy Industry in the future."

"In the past, we pursued machine quality, speed, and efficiency. All of these are visible competition. In the future, the competition in intelligent manufacturing will be intangible. These intangibles are intelligent services based on customer demands and different commercial scenarios. Th[ese] type[s] of intelligent services need to be analyzed and explored from industrial big data," said Jie Li, professor at the University of Cincinnati. Professor Li also pointed out that industrial big data itself is not enough, and that it is only a technical tool. The core objective of industrial big data is to create value, and the value involves avoiding and solving intangible problems and creating new knowledge in the intangible world.

Building industrial big data architecture not only involves hardware technology as sensing, acquisition, and transmission; it also involves the corresponding data analysis and processing software platform. For the enterprise specializing in industrial R&D and production, it is not easy to build a big data application platform.

Cui Peng, industrial market manager at National Instruments (NI), introduced the NI solution to industrial data building. The adopted hardware is a modular CompactRIO embedded controller based on open and flexible LabVIEW RIO architecture; it can perform acquisition and control for different types of sensors, which can realize the data acquisition of different industries, different devices, and different signals. As for software, the online status monitoring suite InsightCM Enterprise, newly promoted by NI, is used for data management, data analysis, and alarm generation. This simplifies the remote management for large-scale deployment of CompactRIO monitoring systems.

As for a big data application algorithm, IMS center led by professor Li Jie cooperated with NI to develop the Watchdog Agent pre-diagnosis toolkit based on LabVIEW. A patented analytical technology and industrial standard are adopted for the toolkit, and the pre-diagnosis and health management algorithm and graphic display function that are easy to use may be added to LabVIEW application; feature extraction, main composition analysis, and pattern matching can be realized, so as to detect and forecast the fault of various objects from key equipment to the whole machine. Meanwhile, it also can help users build user interface, and the forecast result and the analysis information of assets to be tested can be transmitted to customers easily.

Stone Shi is executive editor-in-chief, Control Engineering China. Edited by Joy Chang, digital project manager, Control Engineering, CFE Media, jchang@cfemedia.com.

MORE ADVICE 

Key concepts

  • In the future, the competition in intelligent manufacturing lies in big data.
  • The core objective of industrial big data is to create value and solve intangible problems.

Consider this

What is your plan on building the big industrial data infrastructure? 


Author Bio: Executive editor-in-chief, Control Engineering China