Consumer goods have a life expectancy that ranges from 10 or more years for major appliances and systems to as little as 12-18 months for mobile technologies. Thus, new technology and related standards tend to propagate through consumer goods in relatively short order.
Industrial equipment, in contrast, tends to have much longer lifecycles. Consider an example of a technology used in both consumer and industrial applications—processors.
Processors for consumer products could have a production life span of about five years. However, processors for industrial systems have much longer production runs, and industrial board makers, such as Advantech, take pains to keep any engineering changes to a minimum. This reduces the high cost of design changes, costly maintenance and upgrade efforts.
There are good reasons to take this approach. Industrial systems are subject to greater temperature changes, more vibration, increased dust and other environmental extremes than is the case for consumer applications. These facts require industrial systems to undergo special qualification and careful design to function properly. What’s more, these control and automation systems must often carry out tasks in a manner that can be certified as safe. Once a certification is obtained, there is even more reason not to make changes. For many applications that are only concerned with machine control, a simpler and less powerful processor might be more than enough, as proven by the still widespread use of 8- and 16-bit controllers in industrial applications.
What this means in the context of the Industrial Internet of Things (IIoT) is that many legacy systems were designed a decade or more ago. Though the IIoT was certainly a possibility then, the basics of connectivity, protocols and programming languages were still being worked out. Consequently, industrial system designers frequently opted not to include any communication capabilities. Doing so simplified the design, saving money and increasing security. After all, a system that is not connected cannot be hacked or otherwise compromised externally.
Designs and applications that did include a communication capability typically offered one of several competing approaches that were popular years ago. Whatever the selection was, it constrained the communication channel and capabilities available to the system for the life of its use.
Despite these issues, linking these legacy systems to the IIoT can bring substantial benefits. One is that doing so can help break the barriers that separate operational and information technologies. If the connection is done correctly, the distance between automation and control systems and data analytics packages can be erased. Data can flow from the factory floor, be analyzed, and appropriate actions and adjustments to manufacturing made. Remote access and monitoring of systems in a plant could also be enabled, leading not only to better control but also better maintenance.
Bridging the gap
An IoT PC gateway can overcome the challenge of connecting legacy devices for IIoT applications by linking systems on the factory floor to the cloud. In effect, these gateways supply legacy systems with their missing communication capability. What’s more, since it is a single, separate add-on unit, it can be upgraded and changed out as needed. It’s also possible to install them in stages—deploying them first to those systems that provide the greatest return on investment and then rolling them out to others when doing so makes the most sense.
An IoT PC gateway, like any communication solution, must be cost-effective. This is an important consideration given that they might be installed on or alongside other equipment that is substantially or fully depreciated. Because any addition to an existing system could have a significant impact on the bottom line and profitability, it becomes imperative for the IoT gateway to be cost-effective.
A gateway should also offer wide and comprehensive protocol support. As an add-on, a gateway will have to successfully interface with a variety of programmable logic controllers (PLCs) and other devices, which might communicate via different interfaces and protocols. The gateway should also handle data acquisition and protocol conversion of the data into an appropriate format.
When considering an IoT PC gateway, remember that selecting one with extensive computing capabilities is not always necessary or preferable. This is because heavy-duty analysis can best be done elsewhere, such as in the cloud, where compute power can be added on an ad-hoc basis. Also, an industrial PC gateway can meet the need for higher computing requirements and satisfy several other important parameters at the same time.
Two other points to assess when investigating IoT PC gateways for legacy equipment are compactness and modularity. The need for compactness arises because any gateway will be an add-on to a legacy system. The amount of available space might be very limited, which means that a communication solution should take up as little volume as possible. Given what can be fit into a small and arbitrarily sized space, it might be necessary for a gateway solution to be tailored so that it offers only the bare minimum of functionality. That is easier to do if a gateway has as flexible a form factor and configuration as possible.
Finally, any IoT PC gateway must provide web and cloud access, as well as offer support for a human-machine interface (HMI). The first option is important for any remote access. The second is extremely useful when changes are going to be made locally. Again, a solution based on an IPC can offer such capabilities.
Reaping the benefits
Advantech’s various UNO offerings are examples of such IoT PC gateways. This product family includes X86 systems (UNO-2271G and UNO-2272G) as well as others based on RISC and Quark processors (UNO-1251G and UNO-1252G). Their compact designs support 3G, 4G LTE and low-power WAN connectivity. With Advantech WebAccess/HMI on these gateways, they can support more than 450 types of PLCs and I/O drivers.
Consider, for example, a lathe used to process a metal part or a laser that welds two pieces together. Using one of the Advantech gateways listed above, either machine could tally up how many parts are processed in an hour or a day, how long the operation takes and various other bits of information, such as sensor readings to determine the relative success of material processing.
These data can also be combined with other inputs from machines or systems, either early in the production process or later, such as a final quality control sensor and associated QC checks.
All of this information can then go through analysis to allow, for instance, the spotting of trends. One machine might consistently output product that has a greater likelihood of being in spec and a lower chance of being rejected. A second might do just the opposite. Big Data analytics can reveal such trends, particularly those that involve interaction between machines or conditions that only arise in specific machine processing sequences. The insights possible with this type and volume of data include determining which machine or set of machines makes the best product and offers the highest productivity. Such information, in turn, can lead to better and more streamlined processes, thereby increasing throughput, reducing cost, improving quality and even cutting energy consumption.
Beyond that, more data can also improve machine maintenance. For example, linking information on the status of a system with the quality of its output and analyzing this data can uncover patterns that can be used to predict machine health—even if there is not active machine health monitoring in place. These patterns and the associated data could then lead to proactive maintenance, allowing manufacturers to move from a reactive stance, in which problems are fixed after they happen, to one in which issues are resolved before a machine goes down and product is possibly ruined.
There are many benefits to such a proactive approach. For example, maintenance can be scheduled in advance and at times when the impact on output is minimized. Maintenance processes could also be reduced by fixing machines only when there is a need and not according to a rigid schedule. Finally, the chance that production will be out-of-spec and therefore require either rework or scrapping can be lessened. Together with less unplanned downtime, these benefits can yield a substantial payback.
For more information, visit www.advantech.com.
