There are benefits that edge computing is bringing to the operational technology environment for manufacturers.
Edge computing insights
- Edge computing can significantly streamline and secure operations in industrial environments by processing data locally, reducing reliance on centralized systems.
- Adopting edge computing allows for real-time data analysis and decision-making, enhancing productivity and operational efficiency.
- Integrating edge computing with AI and machine learning technologies helps address labor shortages and enhances cybersecurity measures.
Several of the most important responsibilities for modern control engineers can be made simpler, more autonomous and more secure through the adoption of edge computing. From trouble shooting and maintenance, through optimization, collaboration, compliance and cyber security — the edge environment presents a way of visualizing and processing data that offers potentially huge advantages when compared to less digitalized methods or even the use of cloud data management.
“With the right edge computing solutions, these benefits can be gained with minimal IT expertise at the application level,” said Greg Hookings, director of digital industries – EMEA at Stratus Technologies. “This is vital for a swathe of control engineers who have limited access to ongoing IT support.”
The need for edge computing solutions continues to grow as businesses of all sizes explore artificial intelligence (AI) and machine learning capabilities and deal with ongoing skills and labour shortages, while also protecting their operations from cyber threats.
In the data-rich environment of modern industrial environments, edge computing plays a crucial role in delivering timely insight into application management. “The primary reason for this is down to latency,” Hooking said. “While traditional approaches to data management require collating data for processing in a data centre, or more recently in the cloud environment, this necessarily means that data needs to travel, usually away from the application edge, in order to be processed. This limits the potential for making decisions quickly enough to react to fast-changing conditions of the application. Edge computing offers control engineers much more localised control with data being managed at the application level in real time. This brings forward opportunities to take proactive decisions that can reduce downtime and improve the vital overall equipment effectiveness (OEE) metric at the machine level.”
Hookings pointed to another factor driving the growth of edge Computing – cybersecurity – especially given that controls systems are increasingly targeted in ransomware attacks. “Edge computing platforms can reduce this risk through the use of real-time threat detection monitoring and rapid response to potential ransomware threats,” he said. “With data stored, processed and secured locally, the exposure of control systems to external threats is also reduced, making edge computing a more inherently secure model for managing critical data.”
A paradigm
Daniel Korte, technology manager PLCnext Technology at Phoenix Contact, considers industrial edge computing to be a paradigm that combines the advantages of cloud computing and local data processing to enable more efficient, secure and flexible production systems. “By using open standards, interoperable technologies and real-time communication, industrial edge computing can help the manufacturing industry achieve higher levels of productivity, quality and innovation,” he said.
Korte highlighted the main benefits for industrial edge computing as being lower latency and data protection while obtaining added value from data. This enables cost reductions as well as productivity increases, or improved quality, to be obtained.
“Lower latency is enabled through data processing close to the source, reducing the delay and bandwidth consumption of sending data to the cloud. Data protection can also be achieved as the data is in the local network,” he said.
The specific applications for industrial edge computing, according to Korte, include connectivity and data infrastructure applications. In edge computing, these include the brownfield connection of machines in order to leverage the added value of the data without exchanging the PLC or accessing the logic of the machine. “This is mainly about connectivity with data buffers and data pre-processing,” Korte said. “The data pre-processing, of course, also includes the use of AI, which enables real-time decision making, feedback and control. Preconditions and trends in industrial edge computing are openness as the leading variable and no proprietary technologies, interfaces or protocols should be used. Open Catalog Interface (OCI) containers and virtualisation in general are key technologies here, as a basis for the use of applications that is as hardware independent as possible. In addition, the deployment of these applications must function via standard mechanisms.”
Data management revolution
According to Giacomo Tenerini, hardware developer at Exor International, the growing integration between IIoT and edge computing is revolutionising the management of data generated by industrial devices. “While the cloud approach has traditionally been used, the exponential increase in data necessitates deeper consideration,” he said. “The process of sending data to the cloud for processing can be slow, especially for IIoT applications that require near-real-time responses. Edge computing is a better solution, enabling processing close to the data source, reducing latency and optimizing information flow management.”
Tenerinin highlighted the primary motivation for adopting edge computing as being latency reduction, bandwidth optimization and managing associated costs. Furthermore, he pointed out that the integration of analytics algorithms and machine learning models at the data source further enhances operational efficiency.
Security issues
Dan White, director of technical marketing at Opto 22, highlighted a prevailing trend in industrial edge computing as being a pivot towards enhanced cybersecurity. “Edge manufacturers recognize that robust cybersecurity is not just a protective measure, but is also a key enabler of broader connectivity, both for on-premise systems and especially for cloud-based computing platforms,” he said. “Tools like advanced analytics, anomaly detection, AI, machine learning and large language models, all rely on abundant data, and the edge is where that data resides – in the form of sensors/actuators, industrial equipment and various IIoT devices.”
Edge devices serve as the linchpins in establishing secure connections to cloud services and when considering the right edge device, White advises that engineers should look out for the following features to ensure it is possible to securely democratize data:
-
Integrated firewall management: Built-in firewalls that scrutinize and regulate network traffic are essential for maintaining network integrity and reliability in OT settings. These firewalls filter traffic to block unauthorised access and potential threats.
-
Encryption and security certificates: SSL/TLS encryption ensures secure data transmission and protects against data breaches. Security certificates authenticate communication between devices and systems, allowing only verified exchanges.
-
Network segmentation: Networks should be segmented into secure zones, protecting critical systems from less secure networks and thereby mitigating cyber threats.
-
Zero trust user management: Secure edge devices should avoid default credentials and backdoors for password recovery. Look for devices that do not inherently trust any user or device and necessitate rigorous authentication and authorisation, often managed through Lightweight Directory Access Protocol (LDAP). Strict access control is crucial for preventing unauthorized access and securing OT systems.
-
Pub/sub communication methodologies: Support for advanced publish/subscribe (pub/sub) communication models like MQTT enables efficient and reliable data exchange between devices and systems. This model is particularly effective in environments requiring real-time data distribution. Protocols like SparkplugB, recognised as the international standard ISO/IEC 20237:2023, ensure interoperability among devices, further solidifying the cybersecurity framework within OT environments by enabling secure, structured and efficient data communication.
“The trend towards prioritizing cybersecurity in edge computing is clear and fundamentally about data democratization,” White said. “Whether connecting to an on-premise server or a cloud computing platform, enabling more secure and scalable connectivity ensures that production systems are not only resilient against increasing cyber threats but also primed for future growth and innovation.”
– This originally appeared on Control Engineering Europe.
