Gathering Big Data analytics through network monitoring
Big Data is thriving with multiple levels of information cascading across networks on a global basis with the direct intent to hike business efficiency and productivity while profits remain on an upward trend. When a manufacturer employs Big Data analytics, that means the business is much more dependent on the healthy and correct operation of the network than it ever has been before. The business is making decisions based on those analytics and if that data is not available, it can cost the business big time.
When shifting the business over to a stronger dependency on technology, a network monitoring tool is very important to ensure everything remains functional and available. That’s because networks are getting much bigger and more complex, and as a result, there is a greater potential for multiple points of failure. A monitoring tool that keeps a watchful eye out for any missteps can help mitigate any issues. One case in point is manufacturing chips at Intel. The industry giant has to test every chip that comes off its production line. That means running each chip through 19,000 tests.
Using Big Data for predictive analytics, Intel was able to cut the amount of tests required for quality assurance. Starting at the wafer level, Intel analyzed data from the manufacturing process to focus on specific tests.
The result was a savings of $3 million in manufacturing costs for a single line of Intel Core processors. By expanding Big Data use in chip manufacturing, the company expects to save $30 million.
Keeping an eye on the network
But if the network is not able to pull key information from a myriad of sensors and get it into the proper analytics engine, then forget about the millions of dollars in cost savings.
On top of that, Big Data also now involves the Industrial Internet of Things (IIoT), which means an additional flood of data is coming from more sensors, which also can drive up the sophistication and the size of the network.
Further, in certain processing environments like pharmaceuticals, chemicals, and mining, extreme swings in variability are common.
Given the added complexity of production, manufacturers need a deeper and more intense way to diagnose and correct issues. That is where analytics from a network monitoring tool come into play. By adding statistics and other mathematical tools to business data, it is possible to assess and improve practices.
In manufacturing, operations can use advanced analytics to dive deep into historical process data, identify patterns and relationships among process steps and inputs, and then optimize the factors that prove to have the greatest effect on yield.
In one survey, Tata Consultancy Services asked manufacturers to rate the following Big Data benefits and the biggest were product quality and defects tracking. Other benefits mentioned were supply planning, manufacturing process defect tracking, and supplier, components, and parts defect tracking.
With more users taking advantage of the opportunities analytics bring to the table, their networks are growing quickly. With that growth and reliance, being able to monitor what is going on is worth its weight in gold.
From a Big Data point of view, manufacturers want to see the massive amounts of information they can process from other tools. It is possible to pull data from the network monitoring tool and export or extract that information out into other Big Data pulls that have analytics engines already developed.
It is then possible to view network data and cull any vital facts showing possible anomalies or issues that support growing productivity within the manufacturing enterprise.
In the operations technology (OT) environment, manufacturers are using Big Data to drive improvements to their product and their process. In that world, they have greatly expanded the size and complexity of their network to add these sensors in an effort to boost process knowledge and then drive the data back into their servers.
They must ensure they can communicate with their sensors through their network, so that ends up being a real-time monitoring task.
While the positive side is a vast wealth of knowledge that can help move the manufacturer forward, the flip side is that the move greatly expands its attack surface, so they need to have the capability to capture—and store—vital data to be able to do a forensic investigation just in case there is some type of accidental or malicious incident on the network.
Along those lines, it is possible to archive network data over days, weeks, months, and years, to provide a different insight as to how the network has changed over time. Typically, the user would view capacity levels and then determine if an upgrade is in order. Now, with all the data available, the user can ask, "Do we need to upgrade or do we need to change what we allow on the network or should we split the network and take the traffic with this type of communication and move it over here instead so we are not spending too much money in one area when we don’t have to?"
Big Data comes down to networks and functional business units and the more the industry starts using networks and information technology (IT) resources, the more data that will end up generated which could give a big boost to any manufacturing enterprise. That changeover to a stronger dependency on technology doesn’t have to be difficult. A network monitoring tool can ensure everything remains functional and available—and the process keeps rolling along profitably.
Gregory Hale is the editor and founder of Industrial Safety and Security Source (ISSSource.com), a news and information Website covering safety and security issues in the manufacturing automation sector. This content originally appeared on ISSSource.com. Edited by Chris Vavra, production editor, CFE Media, Control Engineering, firstname.lastname@example.org.
See additional stories from ISSSource about the IIoT linked below.