Living on the edge
Industry's survival depends on moving a significant percentage of processing intelligence away from the data centre and the cloud and closer to operations.
Many industries are under increasing pressure to become more competitive and efficient in order to remain profitable and viable. As part of their digitisation strategies, companies are looking at edge computing in industrial spaces to drive better real-time decisions and possibly even control industrial processes. This has led to much excitement around the industrial Internet of things (IIOT) driving industrial transformation.
To this backdrop, leading technology research and advisory firm for industry and infrastructure, Arc Advisory Group has sounded a stark warning in its latest IIOT report that unless companies are actively investigating investing in intelligence at the edge, they are going to be left behind.
Pieter van der Merwe, Africa and Middle East Availability Solutions Architect at Stratus Technologies, which provides both edge and cloud-based services and solutions for Fortune 500 companies and small and medium-sized enterprises from around the world, agreed that better intelligence and analytics at the edge would improve all aspects of operations, from efficiency to safety.
He said in Stratus's experience, which spans more than three decades: "Complex and remote infrastructure places a burden on a business to ensure there are sufficient skills and human capital available at all times to ensure smooth and safe operations."
According to Arc's market report, this shift in focus to the edge is cross-industry and widespread, meaning businesses are shifting focus to more investment in technology developed to improve operations and reduce latency as they strive for real-time business intelligence at the edge.
In the broader concept of the Internet of things (IOT), the edge is the place where computing happens. In the IOT space, the edge is all devices, machines, sensors and work points that are connected to the Internet.
Van der Merwe said: "Supported by edge solutions and cloud-based services, which include real-time data analytics and artificial intelligence technologies, these systems can generate new sources of data and increased efficiency when complemented by newer, open and secure industrial networking standards.
"These 'edge systems' can collect, analyse, store and forward data generated by devices being monitored at the points of production," said Van der Merwe.
In the traditional digital set-up, data from sensors connected to various "things" was sent through the gateway to a data centre and the cloud, where computing would take place. This brought with it a built-in latency that held back productivity and efficacy.
Gartner has said that by 2022, 50% of all data generated by enterprises will be processed at the edge, or outside of the data centre and cloud. This means that a portion of processing which will influence a multitude of factors related to operations or the course of industrial business will happen at the source, bypassing the need to first be transported to data centres and the cloud. The obvious drop in latency and resultant near real-time analytics and intelligence has massive efficiency and value spin-offs for businesses.
In its report that was conducted with 300 end-users from across the world and various industries, Arc said: "Respondents realise that more complex computing is going to take place at the edge. Better systems will be needed at the edge, but they but that will also need connectivity to the data centre and the cloud." The report found 93% of respondents agreed that a mix of computing at the edge and the cloud will form the basis of industrial automation.
One of the most telling findings from the report was that 92% of end-users from around the world agreed that better computing capabilities at the edge would enable real-time business decisions, and the overwhelming majority of respondents said operational issues were driving their interest in intelligence at the edge. This means they are seeking better efficiency and improved performance of their assets.
While a quarter of respondents said they would still be investigating the benefits of investing in edge computing, the majority will be picking partners, conducting pilots and implementing edge processing infrastructure in their enterprises. Twelve percent said they expected to already be enjoying edge-intelligence results.
Arc concluded in its report: "Remote management will be important due to limited resources and skills sets... If you haven't started thinking about your edge strategy, start now. Your peers already are."
This warning is echoed by Stratus's use case experience around the world. Van der Merwe continued: "It is critical to remove human error from the edge. Our flagship ftServer provides an extremely low-touch predictive management service built on the premise of reliability. Fusing specialised edge devices with this expertise ensures uninterrupted operations.
"In our experience, businesses around the world are definitely moving towards a mix of investment in edge technology which communicates with the data centre and cloud. This seamless operability between devices and the cloud provides a very compelling business case," said Van der Merwe.