CTS looks at how edge computing is becoming central to a new efficiency revolution
In the age of the industrial internet of things (IIoT), the speed of data analysis is key to effective operation. Edge computing accelerates this process, allowing for industrial data analysis to be performed at the point of collection.
Edge computing is the term for when process data is collected, processed and analysed in a local device, as opposed to being transmitted to a centralised system. Supported by local cloud networks and IIoT platforms like GE Digital’s Predix, systems that support edge computing are proving increasingly popular as a means of streamlining the effectiveness of IIoT networks.
“For plant and utility managers, this presents a range of opportunities to not only improve the efficiency of operations, but to also overcome some of the limitations of centralised IIoT networks,” George Walker, managing director of Novotek UK and Ireland says. “In fact, there are the three main ways that edge computing drives value in businesses.”
Walker cites greater operational efficiency, overcoming network latency and bottlenecks and lowering operational costs. “Traditional analysis is undergone by transferring data externally, which can delay decision-making as errors take longer to be found,” he says. “With edge computing capable systems, large parts of the analysis can be carried out by the devices collecting the data.
“The benefits of this are two-fold. For one, this can allow plant managers to access partial deep analysis in real time without waiting on lengthy analysis to be carried out externally. This means action can be taken earlier, streamlining the decision-making process.
“The second benefit is that the IIoT platform, such as GE digitals Predix, can automatically respond to operational data. The system will be able to automatically adjust processes in real-time. In effect, this would allow for a self-correcting system that is able to maximise uptime and reduce the need for manual maintenance.”
Traditionally, data analysis is carried out by having smart sensors send all their data to a remote location where it is analysed and processed. This is data intensive and can create problems if a network is not robust enough. Channelling large amounts can cause network latency, which interrupts working within the plant as there will be a delay with transferring messages that run through the same network.
“This is particularly problematic for applications where a system needs to act rapidly to a problem, such as in an industrial oven control system in a food production plant, where even a temporary dip in the temperature can result in a batch being unsuitable for market,” Walker adds. “In addition to this, the sheer volume of raw data that can be generated in an industrial or utility plant is also likely to cause data bottlenecks in the wider network.
“By using edge computing systems and a machine-learning IIoT platform, systems can respond to changes in real-time to prevent problems, while also having edge computers in place to compress the data and reduce network impact.”
Due to the amount of information being produced, the cost of data storage is becoming a growing concern for companies. Edge computing and its ability to process data without transmitting it lightens the load put on the network.
Processed data is also less substantial than raw data as calculations can be made that allow the raw data to be compressed, thus reducing file sizes. As such, industrial companies are able to make more economical use of their cloud servers. By minimising storage requirements and the number of storage upgrades required, edge computing can allow for a lower overall operating cost.
“It’s clear that there are many benefits to edge computing, both from a financial and operational perspective,” Walker concludes. “Whether a business is still considering adopting IIoT technology or is already making use of such systems, edge computing marks a step forward for businesses looking to streamline processes for efficiency and effectiveness.”