The Industrial Internet of Things (IIoT) is generating massive volume of data. While this could mean potential business opportunities for manufacturers, it also means headaches – lack of storage capacity, overloaded networks and inability to sift through data fast enough to come to the right conclusions.
Edge computing offers a solution to these problems. It provides compute power, memory and even some storage at the edge of the network. This, after all, is where IIoT devices and sensors reside. The presence of processing power on the periphery of the network helps manufacturers to collect and analyze data, detect anomalies and drives decision-making.
According to a recent paper by Frost & Sullivan, “Moving to the Edge: Evaluating the potential benefits of bringing shop-floor automation closer to the cloud,” by Karthik Sundaram and Nandini Natarajan, there are many benefits to a manufacturing environment.
The analysis of data, for example is much faster at the local device level. The alternative is transmitting all sensor and edge device data across the network to the cloud or data center either via satellite link or cabling. Either way, too much time is consumed in sending, processing and interpreting data centrally. As well as latency, this overloads centralized infrastructures and creates bottlenecks. Real-time analysis can only be achieved by edge processing.
How about costs? The more data that has to be transported across a network, the more bandwidth is going to be required. And with data volumes due to explode in the coming years due to the IIoT, a centralized compute strategy could cause networking costs to spiral out of control. Operational or data management costs can be contained by permitted the data to reside within an edge device. This not only minimizes the costs of infrastructure for network transmission, it also simplifies the task (and cost) of subjecting that data to analysis.
One additional benefit deserves mention: the IIoT will generates huge amounts of data, and only a small fraction of it has real value. By sifting through and summarizing data at the edge, basic analysis can aid task automation. And network transmission can be reserved for only most important data. Servers and other smart devices on the edge, then, can reduce the load on a centralized infrastructure and eliminate the need for extensive upgrades to keep pace with an exponential rise in traffic.
Sending IIoT sensor data to and from the cloud or data center, therefore, is time-consuming, expensive, and unpractical. With manufacturing becoming leaner and just-in-time principals becoming the norm, every second matters. Slows caused by data movement equates to loss of money.
Frost & Sullivan recommends that manufacturers start small when implementing edge computing, prove its value and then scale it up broadly. Further tips included:
- Making frequent iterations to align edge computing and associated infrastructure with business processes to ensure it delivers the right insights in a timely manner. Further, as technology evolves and even more processing power becomes available, revisiting edge architectures can provide additional capabilities.
- Train IT personnel on edge and IIoT technologies to maximize their value.
- Ensure device interfaces are securely configured with appropriate access controls and physical security.
- Connect assets to the internet by adding sensors, beginning with systems such as SCADA and MES.
- Invest in edge infrastructure now rather than continuing to deploy traditional architectures which may become obsolete in a year or two due to lack of IIoT functionality, and their inability to provide localized insights.
By implementing edge computing now, organizations are laying a pathway to true digital manufacturing. Once IIoT sensors and devices begin feeding edge computers with data, it becomes possible to establish autonomous decision-making i.e. artificial intelligence or based on certain pre-programmed parameters and rules, routine actions can be taken without human involvement. More major decisions would be referred to the operator.
Breakdowns and hazardous incidents could be greatly reduced due to timely alerts generated by edge devices. By remotely analyzing data at the site, edge can play a vital role in foreseeing disasters and preventing catastrophes. Similarly, predictive maintenance practices would be greatly enhanced by the presence of edge analytics. IIoT sensors would provide monitoring capabilities to enable manufacturers to keep a watchful eye over component condition and the overall health of assets. Alerts would be generated to highlight areas of potential problems. Unscheduled downtime could be greatly reduced as a result.
The industrial automation systems emerging today are on the forefront of the digital transformation. These systems need the adaptability and agility offered by edge computing. By augmenting plant systems with sensors, IIoT-enabled devices and edge computing equipment, a new era of manufacturing is upon us.
Frost & Sullivan noted that companies like Siemens are further along in adopting edge computing than others. Siemens offers a complete portfolio of edge computing devices and automation technology.
For more information, visit: