“Edge” may be the wrong word. Just as wireless and mobile technologies bring devices, data and users together, digitalization is connecting the edges of formerly isolated processes into links in larger network and systems. Here’s how several suppliers are making it happen.
Emerson
“Edge computing is a growing topic because it offers ways to redesign onsite computing nodes on the OT and IT sides, and enhance connectivity and sensing, including wireless and increased mobility,” says Claudio Fayad, VP of technology for Emerson’s process systems and automation business. “Edge computing also means that not all data has to go all the way up to the cloud. It lets users make data available to mobile devices throughout their plants, and monitor what they need wherever without actuating, and complete or assign tasks as needed. This is also where wireless and edge computing can connect, achieve more mobility, and gain more data to drive greater flexibility and efficiency. Mobility, aided by edge computing is especially useful because, even though the COVID-19 pandemic is over, many users are continuing to work remotely.”
Because of all the added connections that edge computing creates, Fayad reports it also needs cybersecurity. Today, this often means using a zero-trust strategy, which doesn’t have to rely on a specific location such as a control room for changing setpoints. “Because users can do more tasks wherever, they need one security model inside the plant, and another security model outside the plant for whoever has access to running the plant and making changes,” explains Fayad. “Zero-trust means the right person is using the right device, in the right location, and with the proper credentials, permissions and software, which are the five pillars of the Cybersecurity and Infrastructure Security Agency’s maturity model. Likewise, our DeltaV Edge software has been extended to integrate zero-trust with our edge-computing environment.”
To select the most appropriate edge-computing solution, Fayad adds that users must decide if they want more data or greater optimization, and use those goals to identify the critical WirelessHART or other sensors they need to deploy. “Once they pick the right components, they must make sure their edge connections are secure, so they can use laptops and tablet PCs in the same way desktop PCs were previously used in control rooms,” says Fayad. “Once they’ve settled on the right measures and secured their edge infrastructure, they also need to determine what security is needed for their Internet, 5G and mobile device connections.”
Honeywell
“There’s been lots of convergence in recent years of security applications and capabilities with mobile devices and software. Computing on the edge and on mobile devices is more secure than ever. It can run on different devices or remotely, and remote connections can also take different forms,” says Praveen Sam, offering management director at Honeywell Connected Enterprise. “Despite these advances, much of the installed bases in many industries are still disparate systems that are still managed individually. They have air gaps between them, use third-party software, and require firewalls and detection.”
For example, Honeywell Forge software lets process-industry users improve reliability and maintenance, and enables field workers to perform managed, mobile inspections and maintenance. To do this, they need comparable data from equipment at different levels, from sensors to control systems to equipment and remote systems, each with different cybersecurity requirements. With advances in zero-trust cybersecurity and intrusion prevention, these capabilities are converging, enabling remote field workers in ways that weren’t possible before.”
Sam reports one of Honeywell’s clients in Oman is drilling gas wells in the desert. This requires numerous device connections and handoffs among local devices and users, who collaborate with offsite experts elsewhere in the Middle East and Europe. This enables seamless remote operations for drilling operations.
Another example is Honeywell Forge Production Intelligence software that coordinates input and participation by multiple users, applications, facilities and suppliers, which need secure connections and contextualized data for analytics across applications. To aid these efforts, Sam adds that Honeywell is using Knowledge Graph technologies that help provide context and establish relationships for data transmitted via multiple protocols.
“If engineers see a process deviation that’s way off, they have to seek its trail of events,” explains Sam. “This means coordinating between multiple data sources and performing diagnostics. For example, when a valve fails, users need to track instruments that were giving information to the network, operator activities, weather conditions and other sources, and look across all of them for a single source of truth. Honeywell Forge can provide that unified view by interconnecting with multiple systems, and accessing disparate data to apply advanced analytics, which aren’t typically available to process engineers or maintenance personnel.”
In addition, Sam reports that Honeywell acquired Compressor Control Corp. (CCC) in June 2023, which provides advanced turbomachinery controls and optimization. By embedding proprietary know-how, Honeywell Forge provides a base for CCC’s operators to manage performance, and prevent surge events or failures. These functions require constant streams of data and analytics, running in real-time, so users can detect problems early.
“Part of this data unification involves adding more equipment-specific know-how that can show users when they have anomalies,” says Sam. “Honeywell Forge is also adding proprietary intellectual property (IP) from its work with CCC, as well as UOP’s expertise. Previously, process applications used different sets of performance models and curves, but they weren’t as integrated and capable as we can make them now. We still have a series of devices and data sources at the edge, but now we can use better information models with them for detecting issues.”
Siemens
One of the best ways to simplify moving field and OT data up to IT levels for analysis is using edge computing such as Siemens Industrial Edge. For instance, our Industrial Information Hub (IIH) software that runs on Industrial Edge Devices lets users build data asset models and maps at the edge, which takes care of all data integration from the shop floor to the cloud. It gathers data through a wide range of connectors, and makes live data access possible on every layer up to the cloud bidirectionally. Built-in security protects IIH against malicious access and allows secure cloud synchronization. These applications run on edge devices, which provide an integrated and secure runtime environment for running high-level, programming language-based, edge applications based on the Docker IT standard for local and powerful data processing and analysis at the automation level. These devices include Siemens’ Simatic industrial PCs (IPC), Unified comfort panels, IoT2050 and Scalence LPE 9413 switches.
“Siemens offers An industrial edge management system to centrally manage these edge devices and Docker-containerized edge apps, which makes them scalable,” says Chris Liu, product marketing manager for Siemens’ Industrial Edge computing platform, “OEE is one use case. Step-chain or cycle time is another use case. Users can run different edge apps on the same or different edge devices to realize these use cases.”
For example, Berlin-based Korsch AG uses Performance Insight, Machine Monitor and Simatic Notifier software in the Industrial Edge platform to monitor the status of components in its pharmaceutical tablet-pressing machines. Performance Insight lets Korsch’s equipment preprocess production data at the edge, where it’s generated before sending it to advanced analytical processes in the cloud. Results and further instructions can then be sent back to the edge infrastructure to optimize workflows, save resources, and improve product quality.
“Our customer service business also invites Industrial Edge’s partner organizations to customers’ facilities to help them identify problems and solutions,” explains Liu. “For instance, Accenture will send a consultant to identify how they can integrate artificial intelligence (AI) on the edge. It offers a one-day Edge AI exploration workshop that will identify three use cases for that customer. Siemens also provides its AI Model Manager software and AI software development kit (SDK) for downloading different AI models from the cloud and deploying them in edge devices.
“Beyond initial AI use cases, these workshops explore the potential and feasibility of AI on overall edge implementations for business, and optimize operations with AI on the edge to improve service quality, predictive maintenance, automation and safety. These can be achieved with software such as AI SDK to wrap and package AI models. Its AI Model Manager software streamlines large-scale model deployment and download models from the cloud, while AI Inference Server software ensures efficient AI operations, and AI Model Monitor software monitors AI models and device metrics.”
