Dell’Oro Group is cutting expectations for the multi-access edge computing (MEC) market, noting it has so far failed to provide compelling use cases to drive enterprise adoption.

The research firm said it expects the market to come in more than 20% lower than previously forecast for 2023. Dave Bolan, research director at Dell’Oro Group, wrote that the shortfall was tied to MEC being overshadowed by an initial operator focus on private network deployments, a lack of applications with “well-defined return on investment,” an initial overabundance of potential providers in the space, including hyperscalers, and possible technology overlap with connection services like Wi-Fi.

“We believe enterprises are evaluating all of these choices before moving forward with more aggressive MEC deployments,” Bolan wrote. “As a result, we are lowering the MEC forecast for 2023 by more than 20% from our previous forecast.”

The Dell’Oro Group insight specifically points to the multi-access edge computing iteration of MEC, which encompasses broader connectivity opportunities beyond the more cellular-centric approach typically associated with mobile edge computing.

In an interview with SDxCentral, Bolan explained that private networks are driving the market, backed by government efforts to free up specific spectrum bands to support deployments. This includes work in the United States tied to the Citizen Broadband Radio Service (CBRS) spectrum that is core to initial private network deployments.

“There's been a lot of emphasis there, but the work that's done there could be leveraged in use cases that develop into what we call MEC, which is an extension of mobile private network, the way we define it,” Bolan said.

A bigger challenge is in crafting a financially viable use case.

For operators, Bolan said they should look toward services like high-bandwidth IoT devices that are leveraging video and computer vision. He explained this includes devices or equipment with embedded cameras that can track or transmit data in real time over a high-speed, low-latency private 5G network to an on-premises cloud service that can tap into machine learning (ML) and artificial intelligence (AI) in support of enterprise application and services.

These efforts are well underway in China, which has a tightly integrated ecosystem and broadly deployed 5G standalone (SA) network cores that support full 5G capabilities, but have so far failed to gain traction in Western markets.

“China has had a chance to develop the ecosystem and start replicating these applications, but what's happening in the Western world is we're getting bogged down into proof-of-concepts that are two to three years long,” Bolan said.

He explained that these proof-of-concepts require enterprises to work through logistical challenges in reconfiguring legacy infrastructure to support these new connection technologies.

This sentiment was echoed by Douglas Bellin, global lead of business development for smart factories and industry 4.0 at Amazon Web Services (AWS), who explained during an interview at the 2022 AWS re:Invent show about a lack of 5G-enabled devices.

Bellin said the average age for machines across manufacturing globally is 25 years old, which means there is not a lot of equipment that can communicate with a 5G radio. For wireless technologies, the age of these devices aligns with when Wi-Fi was still new and the cellular industry was deep in its 2G technology phase.

“It’s how you retrofit that and manage that and build that, and a lot of time that’s the basic that we have to come over,” Bellin said. “It’s not the AI or ML – there’s a lot of that out there, there’s partners who can do that, there’s a lot of already-done things with that. But, how do you get the data to it? Is it 5G could be the backhaul to that?”

Enterprises are also having to deal with multiple choices when it comes to MEC platforms, which further complicates and extends the decision-making process.

“This is a long-term investment and you want to make sure you make the right technological choice,” Bolan said of the process. “I think with the hyperscalers showing interest in this — which they should because they're already supporting the enterprise IT workloads — so now what can they help with [operational technology] workloads? An enterprise’s coders know how to work with that technology, whether it's [Amazon Web Services], Google or Azure, so that has added a layer of evaluation, if you will, and that takes time.”

Operators are attempting to break the deadlock.

AT&T recently teamed with Dell Technologies and VMware to provide an enterprise-focused 5G mobile edge compute package that can power on-premises private wireless services. The AT&T MEC with Dell Apex platform has AT&T providing its cellular and 5G connectivity, Dell its edge processing and physical infrastructure offered through its Apex-as-a-service model, and VMware its telco-focused virtualization platform.

That technology choice also has to include legacy systems like Wi-Fi, which have been the backbone of enterprise networks for decades. Bolan noted that the upcoming Wi-Fi 7 specification “promises a big jump in performance and reliability and security. … And that's something that adds another option when you look at the operating expense, which is minimal with Wi-Fi.”

The Wi-Fi route also brings a level of comfort for enterprise IT teams, which are familiar with how to operate and manage those networks.

“5G’s not simple,” Bolan said. “It's very much more network function–intensive than what they had before. It's a cloud architecture and they've got to learn how to manage that.”

Bolan added that he was still tweaking exactly how much lower the MEC market will come in this year versus expectations, but that he’ll “definitely be lowering it, though China’s share is going to stay a bit higher.”