Private networks come with the promise of enterprise-level control over a required level of quality of service. Mission-critical applications, such as remotely coordinated control over robots and automated guided vehicles in intelligent factories, require very low latencies and high bandwidth. MEC (Multi-access Edge Computing) will be the handmaiden to tie the loose ends in integrating communication and computing, complemented by spectral efficiencies, to achieve service quality goals. The entry of cloud computing companies is accelerating the convergence of private networks, local network intelligence, computing services, and MEC that will bolster the adoption of enterprise applications at the edge.

Mission-critical enterprise applications advance

Applications migrating from pilots to commercialization include 3D applications, XReality, 3D simulations of digital twins, and computer vision-based enterprise facilities monitoring. Unity, a company specializing in photorealistic 3D visualizations valuable for educational AR applications and robots running digital twin simulations, is collaborating with Verizon benefiting from its 5G Edge MEC services, running on its 5G Ultra-Wideband network. Individual enterprise clients run the applications on their private clouds and securely interconnect with public networks for wide-area uses.

Ice Mobility, a supply chain company, used Verizon’s on-site 5G Edge network to monitor the quality of packaging with computer vision. Microsoft brought Azure Stack Edge, managed by Azure, to the premises of Ice Mobility with its private mobile network capabilities to reduce latencies. 5G lowers the latency between the tower and enterprise premises but not the time to move data from the data center to the point-of-use, such as the factory floor, in the enterprise. MEC creates an architecture to bring servers close to the point-of-use to reduce latencies. 

A new generation of applications is emerging as a result of the distributed MEC architecture integrated with distributed and virtualized RAN. For example, Bosch’s MEC-View project provides a “birds’ eye view” of traffic around a block for drivers who are otherwise unable to see around intersections, landmarks, and buildings. Panoptic views captured by overhead streetside cameras are mapped on a screen that drivers can see inside their cars to avoid unsuspected accident risks. None of this would be possible without the low-latency private LTE MEC networks.

Cloud companies bring edge intelligence with private networks

Public cloud companies are rushing to the enterprise edge to “colonize the private, on-premises edge, including the ‘far edge.”

Microsoft – Microsoft acquired in March 2020 mobile core vendor “Affirmed Networks,” to power its 200 edge zones and on-premises Azure private edge nodes with a virtualized and containerized evolved packet core for private networks.

Attabotics – Attabotics, a mobile robotics company for warehouses, illustrates the advantage of using private networks. Previously, Attabotics used a mesh network for gathering telemetry data to monitor the performance of its robots for analysis on the public cloud. A private cloud, at the edge, provided by Microsoft, allowed it to increase bandwidth and lower latency to process the data locally, which stopped robots from bumping into each other and go around.

Google Anthos –

Google’s Anthos for Telecom is bringing its cloud computing to the edge, with a private cloud, to deliver vertical applications in collaboration with independent software vendors. It will focus on financial services, healthcare, and retail. One of its major partners is Siemens which is looking for industrial-scale applications of AI for industrial automation. Automated quality inspection and predictive maintenance are the primary use cases.

Amazon – Amazon’s Wave edge platform is seeking to lower latencies for enterprise applications. For example, Verizon is collaborating with Amazon for a Private 5G Mobile Edge Compute service with AWS Outposts and has Corning’s fiber optic cable factory as a customer. The goal is to have mobile robots monitor work-in-progress at Corning for achieving the rigorous standards of quality in its factory. Amazon Outpost is essentially a Private MEC, a piece of the Amazon cloud at the edge, to run the “sensing-as-a-service” from Gestalt Robotics. Mobile robots look for quality shortfalls with computer vision, and the data is processed at the edge in real-time to prevent faulty parts from becoming parts of final products.

Conclusions

The spadework for building the infrastructure for the development of enterprise applications is well underway at network service providers. It is accelerating in collaboration with public cloud companies that have offerings for private cloud MEC. The intelligence, the edge data centers, and radio networks are all coming together to launch innovative enterprise solutions. The use cases, especially quality monitoring and remote control of moving objects in factories, have a compelling value, and the enterprise is keen to build solutions for them.