Celona and Megh Computing Collaborate to Validate the Value of Computer Vision Applications Running Over New 5G LAN Technology

Tight integration of Celona’s 5G LAN technology with Megh Computing’s intelligent video analytics at the edge lets enterprises to easily enable a new era of innovative use cases

Celona, the pioneer of the industry’s first enterprise 5G LAN solutions, and Megh Computing, the leading real-time streaming analytics platform provider, today announced that they have successfully demonstrated the concurrent operation of intelligent video analytics and private 5G networking within a unified edge computing environment.

The collaboration between the companies opens the door to new use cases within smart factories, cities, warehouses, and schools such as physical distance monitoring, intrusion detection, Personal Protective Equipment (PPE) compliance, and real-time customer intelligence. These scenarios require the use of advanced AI models to drive data insights from the video signals that need to be captured and transferred on a network backbone with predictable service levels – delivered by a Celona 5G LAN.

In addition, the cooperation by Megh and Celona demonstrates how enterprises can radically improve the agility, performance, and security of new digital initiatives by bringing together 5G LAN software controls and vital business applications running within a unified edge computing environment.

By more closely aligning critical business applications with the 5G LAN infrastructure within a single edge compute environment, enterprises can reduce the number of touches it takes to rollout and support new digitization projects, eliminate external latencies between the network and applications and ensure the highest levels of security.

“There is tremendous competitive advantage for companies looking to extract business value by marrying 5G network and application computing at the edge of the enterprise,” said Özer Dondurmacıoğlu, VP of Marketing for Celona. “This is especially true for use cases where outdoor connectivity is a must, and for the ones where wired and Wi-Fi LANs are simply not applicable.”

With the rapid growth of network-dependent applications, organizations are quickly embracing 5G LAN technology to gain greater efficiency for vital business use cases that require more deterministic connectivity when on enterprise wireless.

“With over a billion video cameras expected to be deployed by the end of 2021 and with over 95% of the video never reviewed, video analytics is widely viewed as the killer application at the edge,” said Prahbat K. Gupta, CEO of Megh Computing. “But today most of this video data isn’t being efficiently processed in real time at the edge by automated systems that can be easily deployed by an enterprise. With Celona, we have now proven how a cloud-native 5G LAN solution helps overcome these barriers.”

“Our work with Megh clearly demonstrates the transformative value of 5G LANs and the need for an edgeless enterprise architecture that gives companies flexible deployment options that are dictated by business demands,” said Dariush Afshar, VP of Business Development and Strategic Operations at Celona.

HOW IT ALL WORKED

In converging the technologies, cellular-enabled AXIS video cameras were connected to a Celona 5G LAN, communicating through Celona’s 5G LAN mobile core and Megh Computing’s intelligent Video Analytics Solution (VAS-100).

The Celona 5G LAN mobile core and Megh’s VAS-100 system were deployed in docker and Kubernetes containers, respectively, within the same Amazon Web Services (AWS) SnowBall platform.

Traffic from the AXIS cameras was routed by Celona’s 5G LAN core software to Megh’s VAS-100 application – while providing complete visibility to connected devices over the enterprise network. Celona’s patent-pending MicroSlicing™ technology was used to automatically identify and enforce real-time quality of service levels for Megh’s VAS-100 application traffic, applying strict latency, packet loss and throughput thresholds.

Source: Celona media announcement