5G, AI, edge and digital twins are interconnected technologies that together offer new possibilities for real-time data analysis
5G, edge computing, artificial intelligence/machine learning (AI/ML) and digital twins are interconnected technologies that together offer new possibilities for real-time data analysis, intelligent decision-making and the optimization of various processes and systems. At the Industrial 5G Forum (available here on demand), experts weighed in on how the combination of these technologies is supporting use cases in areas like IoT, smart cities, transportation and industrial automation.
At Raytheon Space and Airborne Systems, explained the space defense solution company’s Digital Factory Manager Crystal Taton, a “solid data foundation” is paramount. But data itself isn’t enough, of course, and so the company has turned to AI and ML for insights and decision making, as well as to 5G, which Taton said is used “to get the data where it to needs to be” in order to drive that decision making.
“5G helps us to remove some of those connectivity and bandwidth challenges, so we can more seamlessly integrate across those tools, the people, the products and the processes,” she explained. “That real-time data can help us find relationships that may not have previously been as well understood, and it can help us see those things both internally in our factories and externally.”
Some of this data processing and subsequent decision making will happen at the edge of the network, shifting the workloads to local compute nodes, which reduces latency of decisions and corresponding actions.
“When you’re talking about millisecond latency, physical distance between the device and the decision process matters,” stated Andrew Keene, senior director of product management at Volt Active Data. He added that edge computing also reduces data transmission costs and network congestion, which is particularly important in industrial IoT environments due to the sheer volume of data being generated. “Having to backhaul all of that across the telco network and out into the cloud somewhere, many are finding that the cloud vendor business models include fairly significant data ingress costs, and so processing much of this raw data at the edge can be cost-effective,” he continued. Processing at the edge can also offer added security because the data is potentially never leaving the enterprise premises.
Keene also referenced smart cities as a good study of how data, 5G and AI/ML are working together: “Major urban areas are using 5G capabilities to enhance their traffic management solutions by enabling proactive traffic routing decisions … with 5G bandwidth and latency, smart cities have the ability to take feeds from hundreds of traffic cameras or other endpoints and to use machine learning and artificial intelligence to predict behavior and spot anomalies as they happen, so traffic can be routed away from problem areas in real-time,” he said.
Also at the session was Kelly Watt, program manager of AI and ML at The Dallas Fort Worth (DFW) International Airport, who provided the digital twin angle. He explained that the key aspect of the digital twin is to be able to take historical, current and IoT data — all of which are typically siloed — and bring them all together, delivering a “spatial context” to enable better decision making.
“We have projects in our terminals, with our passenger boarding bridges, conveyance and elevators, even our runways and the condition of the pavement. There’s a lot of information we need to monitor from weather to sensors on the pavement to just the condition of that pavement, which is being uploaded from assessments, people in the field,” Watt said of the DFW Airport’s need to bring all of its data into a single context.
DFW’s digital twin is in the cloud, Watt shared. “Then, our controls will be left off the cloud in a secure edge-level infrastructure or ecosystem.” However, he continued, these technologies “need to talk to each other,” and that is why an airport-wide 5G network is a “fundamental component” in making all of these applications successful.
Taton’s summarizing remarks also pointed to 5G’s core position among the other technologies: “The data matters to these industrial applications. We need to make it standard, visible, discoverable, consumable, and we need to have access to it. Some of it we don’t need near real-time. But others, the value doesn’t come until that real-time access is available — and 5G can help us get there.”
