There are many benefits of private 5G: Increased network speeds, higher bandwidth, more reliable connectivity and security, lower latency, support over a larger area and a greater number of devices.

Still, the latter — or lack thereof — is holding the technology back, according to some experts. That is, high-quality devices that can effectively support high-demand 5G and edge use cases.

For this reason, NTT Ltd. has inked a new partnership with Qualcomm with a goal to accelerate private 5G adoption.

With the multiyear engagement, the companies will prioritize the development of 5G-enabled devices and seek to advance artificial intelligence (AI) processing capabilities at the edge. Additionally, NTT is unveiling a new subscription device-as-a-service program through its Edge as a Service offering.

As Shahid Ahmed, EVP for new ventures and innovation at NTT, put it: Devices are the “lowest common denominator” for private networks.

“Without devices, you don’t have use cases, without use cases you don’t have a business case, and without a business case you’re not going to see any private network adoption,” he said.

According to IDC, worldwide private LTE/5G wireless infrastructure revenues will reach $8.3 billion by 2026, a significant increase from $1.7 billion in 2021.

Still, as Ahmed noted, it’s getting there that’s the challenge.

The partnership is driven by customer demand, he said, and the intent is to build on Qualcomm’s “vast ecosystem” of technologies and partnerships to build devices that natively support private networks and private 5G.

NTT and Qualcomm plan to deliver 5G-ready devices with Qualcomm’s 5G chipsets and built-in AI models. This will help to enhance AI at the edge and support Industry 4.0 use cases, including push-to-talk devices, augmented reality headsets, computer vision cameras and sensors.

Particularly in edge AI use cases in industrial environments where real-time — or near real-time —functionality and local data processing are key, device availability, variety and interoperability have been some of the biggest issues, according to Ahmed.

Simply put, for edge AI to be successful, “devices have to work with everything,” he said.

Similarly, the interoperability and connectivity required of 5G to support edge environments is often lacking. Ahmed pointed out that even in open RAN environments, interoperability between access points is often “not quite there.”

“It’s plain-old coverage,” he said. Manufacturing facilities, particularly, have many moving parts and Wi-Fi simply isn’t adequate.

Large factories require “tons of Wi-Fi access points,” he said, which creates backhaul of Ethernet and low voltage cabling. With private 5G, organizations can achieve better connectivity and performance with far fewer physical resources.

“It’s an order of magnitude difference in terms of coverage,” he said. “Coverage is the biggest use case we're beginning to see.”

Edge AI accelerates “the ability to make inference decisions at the edge using computation and AI power embedded in devices” — in Qualcomm’s case, its Snapdragon chip semiconductor product line — explained Atul Suri, VP for strategy and analysis at Qualcomm.

“This reduces the amount of data and processing that has to be done in the cloud, thereby creating an even more economic business solution,” he said.

Ahmed pointed out that in traditional manufacturing environments, “transition as well as handoff creates all sorts of challenges.”

But with edge AI it’s all done locally, resulting in “massive transformational opportunities,” he said.

The companies expect the partnership to unlock new use cases. For instance, Ahmed , organizations will be able to virtualize programmable logic controllers (PLCs) — or industrial microprocessor-based tools with programmable memories that store program instructions and various other functions. This could allow them to simulate within local compute environments, perform what-if scenarios, then roll out new instructions to PLCs that immediately take effect.

Or, advanced image recognition could support real-time item counting or verification that workers are wearing (and properly using) personal protective equipment (PPE).

On a larger scale, “by working with Qualcomm Technologies, we will further accelerate demand for private 5G across global industries,” Ahmed said.

Also announced today, NTT is expanding its Edge as a Service offering to include device as a service. The goal is to broaden access to 5G and edge devices and reduce IT maintenance and costs for enterprises, Ahmed explained.

Customers will now be able to access and upgrade 5G and edge devices on a per-use, per-month fee model that doesn’t require them to make large up-front capital investments.

For smaller enterprises looking to use such devices, “the economics are not quite there,” said Ahmed.

The devices, offered through Qualcomm, will include scanners, video cameras, tables and other IoT technologies such as gateways. Contracts with the subscription model will include regular upgrades, maintenance and support, according to NTT.

“The proliferation of 5G-enabled devices is a critical component of shaping a more digital and sustainable future,” Mark Bidinger, president for commercial and industrial segments and channels at NTT customer Schneider Electric, said in a press release. “It forms the backbone of many tech advancements that can improve efficiency and sustainability through efficient resource management and energy conservation, and [it is] pivotal for innovation across various industries.”

The NTT-Qualcomm partnership, he added, “represents a significant step forward in advancing private 5G adoption and meeting the unique demands of the Internet of Things and machine learning.”