Insight and analysis on the data center space from industry thought leaders.

A Powerful Edge for Industrial IoT

With Long Term Evolution (LTE) and the rollout of 5G, it will be possible to flexibly connect thousands of IIoT devices at a site with industrial-strength wireless connectivity.

Industry Perspectives

August 19, 2019

5 Min Read
Data Center Knowledge logo

chrisjones.jpg

chrisjones

Chris Jones is Head of Technology Strategy and Architecture for the Bell Labs Corporate CTO in Nokia.

The internet has been defined by the build-out of an immense amount of global network infrastructure and commensurate cloud data center infrastructure to host the web services and platforms that have come to redefine how we live and work. General human information retrieval (search, web browsing) and communication, e-commerce and entertainment services (audio, video, email, or text messaging) can be served adequately if they meet the human perception limit of about 100ms. This has allowed the cloud data centers to reside thousands of kilometers away from the end-user and still achieve reasonable service performance.

However, we are moving into a new era driven by the transformation of industrial processes and operations, where this architecture will no longer satisfy the intensifying service requirements. The rise of Industrial IoT (IIoT), or Industry 4.0, changes the scope and scale of interconnectedness between the industrial operational technologies (OT) with information and communications technology (ICT).  The future is about coupling these worlds in real-time.  This places new, and far more rigorous, demands on latency and reliability of the overall network and cloud infrastructure in order to drive to new levels of optimization and efficiency.

Recently with Long Term Evolution (LTE), and soon with the rollout of 5G, it will be possible to flexibly connect thousands of IIoT devices at a site with industrial-strength wireless connectivity. The cloud running these advanced IIoT applications can no longer be centralized and must move closer to the “end-user” which is no longer a human, but arrays of sensors and machines working together. This will enable or enhance a breadth of use cases such as closed-loop coordinated machine control, orchestration of automated guided vehicles (AGVs), 360 degree multi-sensory safety operations, video-based remote operation, real-time video analytics, and ‘blanket’ sensing for process automation as well as predictive maintenance. 

This represents a significant shift from today’s cloud architectures, which have largely put a premium on the centralization of cloud compute resources —forcing any real-time operations to be limited to the local devices themselves. The new distributed edge cloud balances the scale and sharing benefits of cloud with the performance needs for localization. For some industries, the preference will be for the edge cloud to be on-premise. This will be especially true for remote locations, such as mines and offshore oil fields, but may also be true for more common locations when the industry can support the cloud resources with their own IT staff and where strict control and security are paramount.  For some industries, however, a common co-location facility close to the industrial campus, including network provider locations, will be a more optimal solution.

The virtualized architecture of modern networks means that these edge cloud resources will also host the network functions that provide the LTE/5G connectivity for IIoT.  Indeed, achieving the end-to-end performance necessary to achieve the most stringent Industry 4.0 use cases will require close integration and optimization between network, cloud, application and end devices.  This makes possible new and hybrid business models for network and cloud services, as well as with IIoT application providers.

This flexibility can play out differently for different use cases. For instance, an oil or mining field camp might set up a local LTE/5G network to provide connectivity and edge cloud support for exploration activities in the area, such as gathering geophysical and geochemical data, near-surface seismic imaging and drone-based aero-magnetic surveys. Much of the processing and analysis can be done locally for faster results and less need for expensive satellite and/or microwave links to distant corporate data centers. If the exploration is successful, this same network can be scaled to meet the automation and monitoring needs of the entire mining operation or oil field.

Manufacturing has long employed robots and automation of various kinds. However, by moving to industrial-strength mobile wireless technologies, the movements of robots, AGVs and smart-tool-equipped workers can be orchestrated to achieve improved workflow efficiency and safety. With edge computing built into the private wireless network, it is also possible to deliver the extreme reliability and low latency response times required by machine automation as well as deliver highly responsive interactive services to workers, such as augmented reality (AR) overlays for equipment repair and maintenance, or safe and efficient remote operation. 

Smart cities can use edge computing capabilities to power IIoT analytics at the edge. For instance, when monitoring roadways and intersections with intelligent highway systems (ITS), edge-powered analytics programs can correlate across all sensor data and then only transfer data and video related to anomalies that might indicate an accident or hazardous condition on the roadway. This dramatically cuts down on the amount of data stored locally and transferred over the network. Crucially, it also means that city employees only monitor footage and data that counts. At the same time, data from on-board vehicle sensors/cameras connected by services providers could be integrated into the data stream to provide additional local situational awareness and coordination.

As we extend reliable low latency network connectivity into the industrial setting, the use cases for edge computing will multiply. It will power low latency services due to its proximity and will help filter the volume of digital data with analytics to reduce transport back to centralized clouds.  With the rise of artificial intelligence and machine learning, the edge cloud is an ideal location for localized training of models on the large volumes of data generated by the industrial systems.  Similarly, edge clouds can host digital twins, offering a real-time representation of the state and properties of the physical device at that location.  Overall, edge clouds will play a powerful enabling role in profitably harnessing the capabilities of IIoT technologies as we move to Industry 4.0.

Opinions expressed in the article above do not necessarily reflect the opinions of Data Center Knowledge and Informa.

Industry Perspectives is a content channel at Data Center Knowledge highlighting thought leadership in the data center arena. See our guidelines and submission process for information on participating.

Subscribe to the Data Center Knowledge Newsletter
Get analysis and expert insight on the latest in data center business and technology delivered to your inbox daily.

You May Also Like