VIEWPOINT

Why Managing 5G Complexity Is Crucial To AI Rollouts In 2022 (Reader Forum)

74views

After a slow start, enterprises are poised to accelerate their AI-on-5G deployments as they race to deliver new applications and services to customers and partners, writes RONNIE VASISHTA.

Back in 2019, the pitch to enterprises for adopting 5G technology seemed relatively simple: super-speedy connections, capable of connecting hundreds of devices with better latency and security than 4G LTE and the promise of new revenue from applications and services.

Going into 2022, network providers are finding an old idiom rings more true: The devil’s in the details.

Pandemic disruptions and wider work from home dynamics aside, enterprises have been slow on the uptake for private 5G networks because of the cost and complexity of setting them up. Unlike Wi-Fi, where IT departments are widely familiar with setting up networks across business campuses and warehouses, configuring and optimizing cellular networks is much more specialized and has traditionally relied upon a few companies and purpose built equipment.

Enterprises also tend to be loath to spend money in a technology’s early days despite the promise of being able to create many applications and services that promise even greater returns.

The good news is that network equipment providers, software vendors and cloud service providers are listening. AI-on-5G is finally poised to make big gains in 2022 as enterprises began receiving enticing software-defined solutions that dramatically lower costs, hybrid cloud-on premises installations  or innovative pay-as-you-go offerings.

About 52 per cent of companies have accelerated their plans to implement AI due to the pandemic, according to a 2021 PWC survey of global businesses..Out of the companies that participated in this survey 86 per cent state that AI will become the new mainstream tech at their companies in upcoming years.

Here are the top three things that will advance AI-on-5G deployments this year:

Flexible, more efficient equipment: AI-on-5G creates a software-defined ecosystem in which you can use off-the-shelf servers that shouldn’t require telecom specific bespoke hardware. These servers work well because they are compatible with existing IT single pane of glass management and orchestration tools. Importantly, they also can be used, through software definition for other IT applications. When using these technologies of AI and 5G together on standard hardware and software orchestration stacks, it will play a vital role for numerous developing applications, paving the way for several other new consumer and commercial business prospects.

The AI-on-5G platform solves many enterprise pain points in a cost-effective and efficient single platform. Because of the flexibility and accessibility of the platform, it democratizes the marketplace, much like the early days of PCs or WiFi.

Enterprises can pick and choose what they want in these application servers such as adding industrial automation, intelligent video analytics, speech technology, adding facial recognition, and, of course, adding a software-defined, 3GPP-compliant 5G vRAN VNF.

Hybrid deployment models: Another important factor to consider is that the 5G workload, as well as the AI/ML application, can be provisioned to be hosted on either an on-premise server or a cloud server or a hybrid cloud solution. In addition, the enterprise may want to retain sovereignty or neutrality of the application and/or data by leveraging either enterprise-owned equipment, telco clouds or hyperscaler cloud services.

The software defined 5G stack is therefore ideally run as a container and can be treated as a cloud native application in order to enable this level of flexibility. It’s clear that this will require no bespoke hardware or custom silicon based accelerators in the 5G stack to truly meet the flexibility and ease of deployment that enterprises require.

Convergence of AI and OT solutions: New edge AI applications are driving the growth of intelligent spaces, including the intelligent factory. These factories use cameras and other sensors for inspection and predictive maintenance. However, detection is just step one; once detected, the appropriate action must be taken.This requires a connection between the AI and software defined 5G applications running on the IT equipment and the monitoring-and-control of the machines, traditionally running on a separate Operational Technology system that manages the assembly lines, robotic arms or pick-and-place machines.

Today, integration between these two systems, IT and OT,  is becoming essential to enable the promise of factory automation over a 5G network that can take advantage of the 5G benefits. This year, expect to see more integration of AI and traditional OT management solutions that simplify the adoption of edge AI in industrial environments.

Promise becomes reality

The promise of 5G opening new opportunities for edge computing is now closer than ever. Key benefits of 5G will include network slicing that allows customers to assign dedicated bandwidth to specific applications, ultra-low latency in non-wired environments, as well as improved security and isolation and increased mobility and coverage.

AI-on-5G will unlock new edge AI use cases. These include “Industry 4.0” use cases such as plant automation, autonomous guided robots, in line monitoring and inspection; automotive systems like toll road and vehicle telemetry applications; as well as smart spaces in retail, cities and supply chain applications. The simplicity of deployment and management of these software defined systems is essential to make this widespread.

Leave a Response