Allow me to introduce a new approach to field device maintenance—one that uses Industrial Internet of Things (IIoT) technologies to achieve predictive maintenance. Not only does it offer a low-cost way of deploying IIoT in plants today, but it also presents a new business opportunity for device vendors.
Intrigued? Then please allow me to explain.
No one doubts the value of predictive maintenance. Identifying potential problems in advance instead of waiting for them to happen is key to optimizing any manufacturing process. But predictive maintenance systems are expensive, aren’t they? From an end-user point of view, it might seem like you have to be IT-oriented and possess big budgets to get involved. And it probably seems like any technology supplier offering predictive maintenance capabilities would need to be a major player with software packages that only large companies could afford. Right?
To better understand why, ask yourself this: Where does most of the data needed for predictive maintenance come from? From our field devices, of course. Given this reality, who would be the most knowledgeable about those devices? Why, the device vendors themselves.
Unlock IIoT data in field devices
Slave device vendors know what parameters contribute to an understanding of a device’s condition and how that information could be used. These parameters include current draw, device temperature, number of cycles completed, speed and changes over time. But this information is mostly locked up within plant-floor devices today.
Obtaining this data is often referred to as data mining. Thanks to next-generation chips, devices can easily be cloud-enabled. And using lightweight communications protocols known by the magic acronyms OPC UA and MQTT, you can transmit that data quickly and efficiently between devices and the cloud over any standard real-time Ethernet protocol.
Adding MQTT and OPC UA is easy with interface chips such as Hilscher’s netX 90, which is a system on a chip (SoC). It even includes standard sensor protocol connectivity and advanced security options to offer a complete and secure bridge between a slave device and a real-time Ethernet network.
Once you’ve made a device cloud-friendly, there remains the issue of how communications between device and cloud will be handled. It might be appealing to get the data direct from the programmable logic controller (PLC), since it is the main source of data in a plant. But that goes against everything we believe IIoT should be for two reasons: 1) The PLC is already committed to doing what it was meant to do, i.e. control the plant; and 2) the PLC should not be overloaded with spurious roles that it was never designed to undertake. In addition, upgrading it to handle the demands of predictive maintenance using IIoT could be disastrously costly.
That’s where an edge gateway comes in—but not just any edge gateway. It must be one that is passive, with no presence on the network as far as the PLC (or any other active devices) are concerned. Plus, it should present no security risk to control devices because it will be isolated from the control functions. Implemented properly, this edge gateway should just sit there gathering data from field devices during network free time and pass that data to the cloud. In some cases, it can even be used to process data locally.
Edge gateways and predictive maintenance
The edge gateway function described above is called data aggregation. For IIoT-based predictive maintenance, the edge gateway becomes a vendor-branded appliance. It can be added to the network at any time and its task is to act as middleman. Many vendors have personalized edge gateways, scanning only for the known MAC addresses of their own devices. These data are then passed on by the edge gateway and fed to the data management layer, which is basically a predictive maintenance app that could be located in the cloud or on premise.
If vendors IIoT-enable their devices as discussed here, and customize their own edge gateway and remote monitoring app, they are in an ideal position to offer a device monitoring service to end users. This takes maintenance responsibility for field devices away from the user and places it in the hands of the individual vendor. This does, of course, require appropriate maintenance strategies to be agreed upon with the end user but, when properly deployed, an IIoT-based approach could relieve the user of a heavy load in terms of costs and staffing. This makes it a service users will be prepared to pay for.
Building off of this base, a cloud-based monitoring app will also be needed to fulfill the approach. And today there are many options, including established supervisory control and data acquisition (SCADA) technologies or even via the creation of a special app. The vendor you select might already offer an asset management package.
IIoT-based predictive maintenance keeps upfront and ongoing costs low for the vendor. Plus, hardware costs are minimal while design costs become one-off issues. Once built, the strategy can cover a complete product family and encompass an entire installed base in the field. For both reasons, costs can be amortized widely.
If I’ve piqued your interest in the current possibilities for IIoT-based predictive maintenance, send me an email (firstname.lastname@example.org) and we can discuss this idea further.
For more information, visit Hilscher North America at www.na.hilscher.com.