Measuring and evaluating what’s going on in production is an important part of the job for manufacturing engineers. The right metrics can predict imminent failures, locate inefficiencies, and create opportunities. Continuous improvement is the goal of every business, and without analytics-driven insights, it’s anybody’s guess where improvements are needed. Analytics are impossible without reliable data but collecting production data and making it useful are two major challenges for today’s manufacturers.
The solution to both of those challenges lies in that confusing buzzword of manufacturing, Industry 4.0. A succinct way to describe the fourth industrial revolution is this: Digital technology has now been around for a while, and Industry is beginning to learn how to unlock its full potential. As a result, the way business is done is transforming. John Kawola, president of additive manufacturing company Ultimaker, agrees: “The digital age has moved into manufacturing and is starting to have a real impact. That’s already happened—whether it’s robotics, tools, sensors or IOT technology that keeps track of everything in an automated way.”
Challenge: Collecting Big Data from Production Equipment
In today’s industrial equipment market, connectivity has become necessary. “For machine builders, it’s no longer a differentiator to have connectivity. You have to have connectivity to be successful in the market,” said Daymon Thompson, product manager at Beckhoff Automation.
That’s a good thing. The downside is that the standards and protocols enabling that connectivity are proliferating. MTConnect, OPC-UA, and dozens of proprietary standards are in use. In addition, nearly every machine builder is seemingly getting into the software game as well, offering proprietary dashboard analytics software to help users view and track machine data. The problem with these differing standards and software tools are that many of the manufacturing operations out there do not only have one brand of equipment. A shop may stick with Okuma for CNC mills, but prefer Doosan for turning centers, for example. On top of that, factory automation is often built by another party altogether, and may not even be equipped with sensors measuring things like motor temperature. In addition, equipment may be located across multiple facilities or even countries.
How can data be collected in a centralized and orderly fashion from such disparate systems? A key part of industry 4.0 is the industrial internet of things (IIoT). The technological path to IIoT began decades ago, with the earliest feedback control systems.Today, third-party IIoT system vendors such as Siemens Mindsphere, PTC Thingworks and GE Predix give customers a platform to connect all their data to. These platforms are designed to intake the different types of data flowing from different equipment and sensors and allows the data to be centralized and organized so that it can be analyzed and monitored. Allen-Bradley Series 9000 Diagnostic Sensors provide both a visual and electrical indication of a dirty lens condition.
Another component of the IIoT is the “things.” Today’s machine builders know connectivity is important, and so do the manufacturers of motors, drives, and instruments. Smart devices can connect to a factory network, sending data directly to the IIoT platform. For example, a smart sensor such as the Allen-Bradley Series 9000 Diagnostic Sensors can detect when its lens is collecting dust, providing a visual and electrical indication. This data can enable maintenance personnel to clean the lens before the sensor fails, preventing downtime. “Big data doesn’t help anyone,” said Ramona Schindler, business development manager at Siemens Mindsphere. “You have to make big data smart.”
Making big data smart is challenging. In general, machine data can be grouped by frequency. A machine sending low-frequency data may report information such as power state and which program is running. Machines can also send high-frequency data. For example, a laser-sintering metal 3D printer may report sensor data on laser melt pool temperature every tenth of a second.
One issue that arises is often known as “data inundation”: You’ve generated a haystack of normal condition data, and now you have to find the needle that signals an abnormal condition.Once data is centralized on a platform, it can be imported into analytics applications that run on the platform, keeping everything in one place. Some platforms support applications developed by customers, for those who want to develop a proprietary tool. Most vendors offer apps built for different types of data.
To give a specific example, Siemens MindSphere offers an app called Manage MyMachines, which runs on the platform, allowing users to display and manage a global fleet of machines and execute and view analytics on that data.
Industry 4.0 is enabling better access to the valuable data production processes generate. With better access to data, manufacturing professionals can analyze, predict and manage production more effectively and profitably.
Read the full article in Engineering.com