Following our recent webinar with LNS, “Build an Analytics Anywhere Strategy today with the IIoT”, we have received a number of questions around the cost implications of implementing an IIoT strategy and how cloud and edge computing both play a part in an IIoT framework.
While both cloud and the edge are important components of a full IIoT implementation, cost and ROI help determine the degree to which each plays a part.
About three or so years ago, sending data to the cloud was all the rage. New equipment came with embedded sensors, older equipment would be retro-fitted and a combination of Wi-Fi and mobile technology (LTE) would transmit all this sensor data to the cloud. Once in the cloud, one of the expanding list of cloud-based machine learning engines would deliver actionable insights into the data and enable various optimizations – streamlined production, increased quality, improved maintenance, etc. While this looks like an easy way to go, largely bypassing existing infrastructure, there are several drawbacks to this approach.
While bypassing existing infrastructure and automation systems looks attractive, problems arise as you look to fully roll out an IIoT implementation. Not only can the costs ramp up in the cloud for the analytics piece, but the cost of wireless connectivity can also escalate rapidly. There are ways to mitigate the connectivity aspects by using an internal infrastructure with a gateway, but almost by definition, this starts to introduce more of an edge infrastructure. The elastic nature and pay-by-usage model of the cloud can mean that costs vary over time, although the opportunity to shift to a more OPEX type model rather than a traditional CAPEX model may yield some benefits.
The other gorilla in the room is the one that worries many operational technologists in the automation industry; cybersecurity. Transmitting data outside of the plant remains one of the most challenging aspects of digitalization in the industrial automation world. You only have to look at the TRISIS/TRITON/HatMan malware uncovered safety systems in late 2017 to realize the truly insidious nature of cyber threats. Cost concerns, response times and cybersecurity fears are the main reasons many operational technologists are focusing increased attention on edge computing. The concept of enabling analytics closer to plant floor and augmenting control cabinets with more application capability is increasingly attractive. These systems can have a wide range of applications and can be delivered at attractive price points. They can support a wide range of applications, including small scale, localized SCADA, HMIs, Historians and thin clients. Additionally, they can collect and analyze data from localized sensors and control points, before forwarding it upstream to control systems, Asset Performance Management (APM) systems and/or the cloud. Some of this data could be used for real analytics to improve plant floor operations, such as advanced process control, quality inspections and predictive device failure.
Of course, the closer compute capabilities and applications are placed to the plant floor, the further they are from support and skilled expertise to address problems. This is not the realm of IT support and it may be challenging for automation or operational technologists to respond quickly. Consequently, systems that are simple to install, self-protecting and easily serviced by local plant floor personnel offer distinct value in this edge-based environment. This includes the capability for internal predictive maintenance and remote monitoring so that a proactive approach can be taken to servicing these platforms vs. a reactive one, where production has already been compromised and OEE has decreased.
Off-the-shelf IT grade PCs and industrial form factor ruggedized PCs have all been used in these environments for many years. However, as the importance of the edge grows and moving automation tasks closer to the plant floor along with new analytics capabilities takes hold, the reliance on edge systems increases. This is why Stratus has introduced its zero-touch compute platform, the ztC Edge, specifically to meet the needs of delivering more applications at the edge in a simple, reliable and easy to manage solution.