But wait, there’s more—that IIoT can do. Beyond its accelerating talent for getting data where it needs to go, IIoT can provide similar benefits in several other areas.
“Just as its name implies, IIoT is things on the Internet, like today’s houses with HVAC, garage doors, fridges, ovens, washer/dryers and other appliances with smart devices on an Internet protocol (IP) network. However, the real magic happens when they’re all linked to software apps on a smart phone or PC, which communicate, check their status and can send commands,” says John Clemons, MES solutions consultant at Maverick Technologies, a Rockwell Automation company. “We’ve had PLCs networked with Ethernet adapters since the 1980s, and then on the Internet in the 1990s with local area networks (LAN) and added ports, routers and connectors. We gained some security with virtual LANs (VLAN), which was important when networks were segmented between plant and business levels. Now, even devices that aren’t Internet-enabled can get connected with an app via wireless or cellular on a phone. IIoT is also simpler than before now, so a smart valve can put an entire machine on the Internet, and users can control settings like start/stop and temperatures.”
Process and discrete differences
Clemons reports that Maverick implements IIoT for two main types of clients, including discrete industry manufacturers with machining operations and welding robots, and process owner-operators with mission-critical oil field and similar applications. “IIoT lets all of a user’s machines talk to them, manage tasks like SKUs for CNCs, run parts and report back. Process operations must do what’s right to stay safe, so we can also download settings to their equipment, and get back information.
The Industrial Internet of Things is surpassing itself with added data sources, more detailed information and greater insights—if users are open-minded and flexible enough to try it. Read more of this series here.
“For instance, users may still employ stress testing and x-rays to check on weld quality, but they can also use IIoT and analytics to plot and compare tests with previous results, and establish benchmarks more easily. Because so much of IIoT is used for collecting data in close to real-time and analyzing it, we’ve been working with PTC’s ThingWorx software. In the welding example, users often have upper and lower specification limits, and these can be correlated with actual failures observed during testing, if the data is available soon enough. Users may also produce items within specifications that still fail too often, which means they need to readjust their specifications. IIoT can also show when tests become meaningless because nothing ever fails, prompting further adjustments.”
Much better batches
Similar issues face process industry clients, such as those who make batches of flavorings, run tests, and also usually have to wait hours for pass/fail results before they can make adjustments. “This procedure is almost trial-and-error, so it needs to be smarter,” says Clemons. “Users do detailed measurements of what goes into a batch and variances in ingredient conditions, and then do multivariable analyses for multiple ingredients. IIoT components on batch tanks and data from their controllers can tell users almost immediately if about 90% of their batches are OK or likely to need rework, which means they no longer need to wait for hours. They can also track and trace each ingredient in each batch, match them to certificates of analyses, and run predictive models to determine likely batch characteristics, also without waiting several hours. Users still do traditional tests to verify the performance of their models, while the models can support what the tests can be expected to find.”
Clemons adds that IIoT devices on tanks, controller data and predictive analytics further assist batch applications because users can:
- Accept more varied raw materials and be sure their batches won’t be adversely affected;
- Reject materials more quickly that aren’t usable and likely to fail, as well as alert suppliers and hold them accountable sooner; and
- Achieve and maintain tighter controls on batches, which can also allow them to finish faster.
“For instance, we work with a client that processes meat with spices, and traditionally tumbles its batches for 45 minutes, even though the ingredients are sufficiently incorporated in 30 minutes. This is indicated by viscosity changes in the product, which can be measured by torque changes on the mixer, but the client usually ran it longer just to be sure,” explains Clemons. “This data can also be collected via IIoT, so we did a graph, and found torque plateauing at 30-35. This proved they didn’t need to run for 45 minutes, so they began stopping two or three minutes after the viscosity plateaued, saved about 10 minutes per batch, and were able get five or six more batches per day on one mixer. This shows how using real-time data and analytics in conjunction with testing and historian data can make a difference in predicting batches, as well as enabling troubleshooting, machine learning (ML) and condition-based maintenance.
Taking on track-and-trace
In fact, Clemons stresses that IIoT and analytics projects should start with collecting existing data into a historian, making it available to everyone who needs it, and making sure the process operators know the fundamental track-and-trace principles. They used to be unique to pharmaceutical applications, but they were eventually adopted by food and beverage processors. Most recently they’ve been applied by oil and gas companies and even automakers and Tier 1-3 suppliers producing electric vehicles (EV).
“We have an aluminum manufacturing client that sells body panels to EV companies, and they told us they can’t sell to their customers without full genealogies for their products, including aluminum costs and extruding information,” adds Clemons. “Track-and-trace covers all of this, including what’s needed, who worked on it, which machines and equipment, what codes, what process, and when is maintenance needed? IIoT answers these questions faster from all sources, matches them with the right order number and product, and also makes it easier to integrate identification technologies like barcodes, RFID tags and QR codes.”
