Shared from the 12/1/2016 Smart Industry eEdition


Okuma, GE get practical with Industrial IoT

Machining and turbine manufacturing applications rely on Predix, Brilliant Factory to streamline data delivery and performance

Because so much of the Industrial Internet of Things (IIoT) happens in the hazy world of digital ones and zeros running on computers and servers located who knows where, it can be hard to accept that the IIoT can carry the weight of physically challenging jobs in the real world. To help manufacturers and other likely users get past this preconception, two heavy-industry experts showed how their organizations are embracing IIoT to achieve unprecedented gains in performance and productivity.

“We were just at the International Manufacturing Technology Show (IMTS) in Chicago—which features all variety of metal-cutting machines—and people were talking about IIoT in every exhibit and hall,” said Jim Kosmala, vice president of engineering for machine-tool maker Okuma. “Many are starting small and taking baby steps, but there is a big focus on the IIoT because it’s about getting connected and using data to make better decisions. Okuma, for example, is using GE’s Predix as an embedded IIoT solution.”

Kosmala and Andy Henderson, industry analyst, GE Digital, presented “IIoT—Practical Implementation” at September’s Smart Industry 2016 conference in Chicago.


Kosmala reported that many sensors have been added to machine tools in recent years to monitor vibration and other parameters, and they’ve become great sources of raw data. Most are connected by traditional RS232 serial links, fieldbus networks and other industry-specific protocols, or by the more recently developed Bluetooth wireless and MTConnect, which enables users to link and write to plant-floor devices.

“These sensors and links are also good for big data analytics, and can jibe with parts-quality data and operational effectiveness, and be used to discover other useful trends,” said Kosmala. “For instance, the average utilization time of spindles on machining centers is about 65% or about five hours in eight-hour shift, which is ineffective, so we’re trying to get closer to seven hours of availability.

“Because Okuma’s machine control also uses a dualcore, PC-style processor, connecting a machine to IIoT is like plugging a home PC to an Ethernet network. Now, the most elementary use of IoT is on machines in plant-floor layouts with red/yellow/green displays for job scheduling and parts to be cut next. IIoT connections can help machines and users prepare the next job while the present one is running, which reduces idle time, and gets us closer to that seven hours. Plus, we can use these connections to monitor machine health more closely, and perform predictive maintenance.”


Kosmala added that Okuma even developed and launched its own two-year-old app store, which includes many software applications written by contributors. They include the MTConnect Display open-source app, MTConnect Display Mobile for Android users, ABB Robot Studio, Blum Gauging Guide, Coolant Monitor, Decam PartMaker Viewer and many others. Many of these benefit from IIoT connections and help users gain productivity benefits.

“The second most downloaded app on our site lets users stream XML data, and easily see all MTConnect devices that are running,” added Kosmala. “Unfortunately, only 10% of industrial machines are connected like this due to education, financing and security issues, so our number one job is to make these tools easier and more affordable to use, and keep them secure by allowing only read-only data streaming from their Ethernet ports and not allowing any data in.”

Likewise, as part of its cooperation with GE Digital, Okuma is using MTConnect to link its machine controls with the Predix Machine run-time app, which is available in an embedded version or in a Predix Field Agent that connects to a PC data server. Both formats also use OPC UA to send event-based data to GE Digital’s Predix Cloud service and higher-level Predix Client apps.

“These applications can be set up in a few minutes, run active on each machine, and access the cloud as needed. They don’t need data polling that a box demands, and don’t require the hours it can take to go through a regular IT infrastructure,” added Kosmala. “In the future, once users get some of this data, they’re going to want more, and get more out of it. They’ll want monitoring services like predictive maintenance, uptime reports and operations reports, and they’ll want continuous machine-tool improvement from parameters like load-analysis, vibration, runtime, thermal monitoring, overall equipment efficiency (OEE) and energy consumption. By plugging into IIoT and the cloud they can do all they’ve thought of. For machine builders, we can offer alerts on maintenance that will be needed soon.”


Kosmala reported these apps and Internet-based tools can work with Okuma’s 10-year-old ThINC (The Intelligent Numeric Control) ecosystem and partnership program, which now has more than 50 members. ThINC can combine technologies, such as modeling CAM systems, aiding intelligent tools and fixtures, measuring parts and adjusting machines, controlling adaptive cutting speeds, automating robots and bar feeders, and monitoring machine processes and health. But its data can also go the IIoT and the cloud for big data analytics.

Henderson added that GE also builds similar ecosystems around its Predix platform, such as its GE Brilliant Manufacturing Suite, which helps monitor and optimize operations at its gas turbine manufacturing plant and repair facility in Greenville, S.C. Predix and Brilliant Manufacturing technologies monitor and help manage machine health, process optimization, consumable resources and energy use at the plant.


“Our number one job is to make these tools easier and more affordable to use.” Okuma’s Jim Kosmala explained how the company is equipping its machine tools to participate in the Industrial IoT.

“These tools take data and information, and turn them into action and dollars,” Henderson added. “For instance, Brilliant Manufacturing recently prevented a catastrophic failure on the 1,000-hp motor on one of Greenville’s two main air compressors, which avoided the need for a new motors and saved several hundred thousand of dollars in overhaul costs. In addition, high argon flow during a plant shutdown led to a search for leaks, and data from Brilliant Manufacturing identified a 200 cubic-feet per hour leak in the coating furnace. This prevented unplanned downtime costs and saved thousands of dollars in wasted gas.

“Finally, some of Greenvile’s machines weren’t completing full cycles, were producing defects and recuts, and were idling too long. As a result, Brilliant Manufacturing tools for shift-by-shift utilization and completed operations, live feeds of machinery activity, and defect and work-order tracking contributed to improved tool utilization, enhanced fixture and part loading, and standardized operator work schedules.”

Henderson added the Greenville plant uses Okuma’s MB5000H machine with spindle loading, so the two firms recently collaborated to use data from its thermal and vibration sensors to help its computer numerical controls (CNCs) calculate thermal errors and independently compensate its axes. This allowed MB5000H to migrate from unsupervised operations and clustered data to use machine learning principles to manage its spindle loading and automatically identify anomalies.

“Looking at the loads, wear, broken tools and abnormal events on spindles, and analyzing their data means we can predict cutting, spindle and bearing issues ahead of time,” he added. “Predix, Brilliant Manufacturing and IIoT let us take this data, create a proof of concept, identify performance that’s normal or not, and find trends sooner.”

See this article in the e-Edition Here