4 ways to bring legacy manufacturing equipment into the IoT age
Until recently, connecting manufacturing systems to the Internet of Things was a complicated affair. Older legacy equipment probably dominated the plant floor, unless you worked at a high-tech manufacturing facility. The older manufacturing equipment included manually operated machinery, outdated PLCs that control assembly lines and other manufacturing processes, stand-alone CNCs that control machine tools such as lathes and grinders, and older automation systems. None of those systems were designed to work with modern IoT systems. As a result, bringing a manufacturing plant into the IoT age required expensive upgrades and custom development work.
Today, connecting legacy operations to the IoT need not be so challenging. Falling costs and flexible new technologies have transformed the conversation about integrating legacy equipment. In many cases, what was once assumed to be difficult or impossible can not only be accomplished, but it can be done far more easily and for less money.
"In the past, folks who wanted to get real-time data from their CNC machines had to deal with the high cost of upgrading their controllers and putting in a network and so forth," says Richard Phillips, director of the smart manufacturing and digital transformation group at systems integrator Polytron in Duluth, Georgia. "Now, more and more, there are other solutions out there for them."
Declining costs and ease of installation open up IoT to a wider range of manufacturers and allows for low-cost pilots to test ideas or establish proof of value, Phillips says.
"Now, anybody and everybody can afford that solution and will very likely get a lot of value out of it," says Phillips. "It's a good starting point to prove to themselves that for their manufacturing operations, there's value in deploying IoT solutions."
Why legacy equipment is here to stay
Legacy equipment is deeply rooted in the manufacturing world. Except for state-of-the-art manufacturing plants built in the past five years, most facilities continue to depend on production equipment and other systems that may be decades old.
"Legacy is equipment that has not been defined with connectivity in mind, other than maybe to the local control room or local dashboard for the operator," says Wolfgang Decker, a managing director for digital industry X.0 at Accenture. He says that on average, industrial equipment is 20 to 25 years old, but some equipment will be far older. Decker recalls meeting with a vendor of hydropower pumps who told him they still have 50-year-old pumps out in the field and under maintenance contracts.
This may come as a shock to IT managers. In the data center, hardware and software upgrade cycles take place every few years to accommodate ever-more powerful applications, storage and networking capabilities, and security requirements. Complete rip-and-replace IT overhauls are not uncommon.
On the plant floor, replacing older manufacturing equipment comes with a different set of considerations. Legacy equipment might represent millions in capital expenditures and years of planning, not to mention significant investments in supply chain management, process improvement, safety systems, and operator training. Equipment is purchased with the knowledge that it will last decades and can be depreciated over its useful life.
However, manufacturers have begun to appreciate how Industrial IoT (IIoT) can transform operations, give insights into performance and quality levels, and even create new lines of business. It’s now possible to automate and optimize processes in ways that were inconceivable 15 years ago, using powerful computing hardware and new types of sensors located on the plant floor.
The current challenge is how to more cost effectively implement IoT architectures that include legacy equipment. The traditional legacy-to-IoT path has involved vendor-supported upgrades and custom integration, which may not be feasible. Upgrades may not be available, costs are too great, or the solutions do not scale. However, the following technologies may provide a different path to IIoT:
- Smart sensors: Flexible, low-cost IoT sensors and sensor packs that can be quickly deployed in industrial scenarios.
- Retrofit kits: Devices that bolt onto certain types of equipment and bring sensors, connectivity, control, and other features.
- Edge gateways: A new generation of edge gateways that includes advanced features to support the integration of legacy machinery.
- Video cameras: Decreasing cost, higher resolution cameras, and enhanced analytics that can monitor worker location for safety, read analog gauges, or capture remote equipment status (either thermally or observationally).
These technologies are not a panacea, but they do offer lower cost alternatives and new capabilities in industrial settings.
Using smart sensors to get real-time insights
For companies working in a legacy manufacturing environment, a significant concern is lack of real-time visibility into operations, says Polytron’s Phillips. Even if a legacy system can generate data, the reports often arrive days or weeks later—sometimes too late to do anything about a problem. Newer IoT technologies capable of delivering real-time visibility can deliver significant benefits, he says.
Ideally, that real-time visibility gives the manufacturer data that helps the company understand the source of performance losses (such as speed, availability, or quality). Armed with that information, the company can take the appropriate action to improve the process, Phillips says. Even a small change makes a difference. "Something as simple as a 1 percent increase has significant dollars attached to it."
Historically, Polytron worked with manufacturers to implement full-blown networks and IoT architectures to integrate PLCs and older equipment. Once connected, managers would have access to real-time data and other insights, which could help improve operations or identify problems along the line. But it was expensive and time-consuming.
What has changed, Phillips says, is the availability of new smart sensors that make it possible to gather real-time data at a much lower price. While not as fully featured as a plantwide IoT architecture, smart sensors offer a low-cost alternative to launching IoT pilots or gathering data from a single process.
Smart sensors can send data to on-premises servers, edge devices, or the cloud. Depending on the application, smart sensors can measure environmental conditions, provide location information, or track a component's progress through an assembly line, among other tasks. IoT platforms can be used to configure the sensors and create basic reports. Equipped with mesh networks that can connect to cloud services and powered by battery or low-voltage power connections, just a few smart sensors can deliver real-time insights, Phillips says.
"It's not part of a wireless network that the plant has to put in. The sensors communicate to each other, similar to home automation," Phillips says. "For a battery-operated mesh sensor, you can mount it to the line. You don't have to do any wiring."
He cites an example of a bottling plant where the managers are struggling to understand why output at a filling line is lagging. Smart photoelectric sensors could be placed at various points along the line—one to count bottles going into the filler, another to count bottles coming out of the filler, and a third to tally cases at the end of line. Data streaming off the sensors could identify which process is slowing down the line or throwing errors.
"That gives them enough information to know where their losses are," Phillips says. He adds that a more sophisticated implementation could take data directly from controllers to generate alerts, performance reports, and more detailed metrics.
Taking the smart sensor concept a step further are OEM sensor packs. These offer advanced analytical capabilities including predictive analytics and even video analytics on the factory floor.
Like basic smart sensors, the sensor packs have the advantage of being easy to install. They wirelessly connect to existing industrial and IT networks without disrupting plant operations. However, the types and amounts of data sensor packs can generate require more robust processing capabilities at the edge. This can take place at the device level or by leveraging dedicated edge hardware to process the inputs and deliver insights in real time.
Retrofit kits for legacy machinery
Retrofit kits are another low-cost technology that enables legacy equipment to work with IoT sensors, advanced analytics, and secure IoT architectures. Some retrofit kits are available through vendors or third-party suppliers for specific types of machines, but others work with a variety of legacy machinery.
Andrew Dugenske, director of the Georgia Tech Manufacturing Institute’s Factory Information Systems Center, told attendees at the IoT for Manufacturing Workshop at Georgia Tech about the retrofit kit it helped develop for legacy machine sensing. The kits can be quickly affixed to older machinery and connected via wired or wireless networks, allowing for better monitoring of plant assets and powering new types of applications that would otherwise be unavailable. He says the kits also help address the scalability problem that plagues custom integration projects.
"Hiring a lot of expensive programmers and DBAs is not going to be sustainable when it comes to scalable IoT," Dugenske said.
One manufacturer is using retrofit kits and Georgia Tech’s digital decoupled architecture to bring its extensive collection of legacy equipment, including manual machines, into the IoT era. According to a maintenance technician at the facility, the goals of retrofitting much of its equipment include predictive maintenance and identifying opportunities to increase efficiency.
"We want to know what is going on with our machine tools," he said during a presentation at the Georgia Tech workshop. "Do we need two identical tools running next to each other if one might be sufficient?"
A critical issue for the manufacturer is identifying protocols and hardware that can work with different types of equipment.
"No two machines are ever alike," remarked the technician, noting that the same model might run different versions of an operating system or have different features. He estimated that the company spends approximately $1,500 to retrofit a piece of legacy machinery with an input/output box. Sensors to monitor motors and measure vibration and other conditions on a piece of legacy machinery cost an additional $100 apiece, he said.
Edge gateways add new legacy capabilities
A third approach to integrating legacy equipment involves edge hardware. In recent years, as the use of IIoT has expanded, gateway devices have served a crucial role in connecting OT and IT systems. Installed on the factory floor and outfitted with both industrial and IT interfaces, gateways can perform a number of functions, including translating protocols and securely connecting equipment, sensors, databases, and applications. This includes older machinery and systems, says Steve Hoffenberg, director of industry analysis, IoT, and embedded technology at VDC Research.
"If a factory has lots of legacy equipment that was developed pre-Internet or pre-IoT connectivity and wants to now connect it to the Internet, the most common way to address that is through the use of a gateway device," he says. "It acts as a kind of intermediary between the different endpoint devices and other elements of a system."
Hoffenberg notes that a new generation of gateways brings additional capabilities, including security features.
"Intelligent gateways are increasingly being used as a way to try to bridge legacy systems into more modern Internet-connected systems," Hoffenberg says. "They can also monitor those communications for security-related anomalies or various kinds of known or unknown malware that might be trying to make incursions into the system."
Vendors are striving to make modern edge hardware and IoT systems easier to configure. A number of partnerships that include hardware, software, and platform vendors have launched turnkey solutions that can quickly discover industrial devices, including legacy machinery. While the turnkey solutions may lower the barriers to entry for businesses interested in launching an IoT pilot or expanding an existing implementation, many companies still need expert help to plan, design, and implement such solutions on the plant floor.
At the end of the day, Accenture’s Decker says the potential value should drive decisions about whether or not to integrate legacy equipment. "Always have the value question in mind," he says. "What do we need out of a particular piece of legacy equipment?"
Decker gives the example of a manufacturer trying to optimize overall equipment effectiveness (OEE), a metric that helps businesses measure productivity. "You should not connect [a piece of legacy equipment] just for the sake of connecting," he says. "But in order to measure OEE for a production line end to end, it does not make sense if you have a dead spot in between."
Video cameras for image detection
Video technology has reached an inflection point. Higher resolution cameras come at a lower cost with increased functionality, making image detection and processing in the camera itself a reality. While processing images at the device level is a step forward, it still requires a "video babysitter" to monitor the images and determine whether anything out of the ordinary is happening.
Another alternative is to "stream the data back to an edge compute device for processing and early analytics," says Peter Moser, IoT and AI strategist at Hewlett Packard Enterprise. In this way, "only anomalies that get detected are flagged and sent for further analysis." Anomalies could include people in restricted areas who represent a risk, equipment that is overheating, or containers that are about to overflow. By analyzing on site at the edge, assessment and response time can be greatly improved, even to near real time.
Edge computing opens compelling possibilities, when you consider the relative ease of integrating video data, data from historians, and new sensor data. Moser reflects, "There’s a bridge between the old OT world and the new digital world that allows you to modernize your legacy environment without disrupting your existing environment."
IoT updates for legacy manufacturing: Lessons for leaders
- Integrating legacy manufacturing equipment into modern IoT architectures used to require expensive upgrades or custom workarounds. Retrofit kits, smart sensors, and intelligent gateways can reduce costs and make it easier to get started with IoT.
- When planning an IoT pilot in your legacy environment, don’t start with the technology challenge. Start with the business problem, and from there identify what pieces of equipment may need to be brought into the IoT age.
- Real-time insights enabled by IoT can yield real results. Unlike IT, where we are accustomed to seeing ten- or even hundred-fold improvements, even a 1 percent improvement has significant dollars attached to it that drop straight to the profit line.
This article/content was written by the individual writer identified and does not necessarily reflect the view of Hewlett Packard Enterprise Company.