SAN FRANCISCO—Whether fog computing is driving 5G or 5G is driving fog, the two are certainly entwined here at the second Fog World Congress, which is celebrating all things fog.
The OpenFog Consortium, which turns three years old next month, has grown from a few founding members—ARM, Cisco, Dell, Intel, Microsoft and the Princeton University Edge Laboratory—to 55 members in North America, Europe and Asia.
Matt Vasey, director, Artificial Intelligence Business Development at Microsoft, was one of those who hosted the first formal meeting of the consortium on the Microsoft campus in Redmond, Washington, almost three years ago to this day. The consortium was announced on Nov. 17, 2015.
“Now we have a huge breadth of partners, from operators, hardware partners, a tremendous amount of software partners that are really thinking about how you build out this continuum of intelligence from the cloud all the way down to the edge,” Vasey told FierceWirelessTech on the sidelines of Fog World Congress.
Plenty of companies, as well as standards bodies and government entities like NIST, are talking about fog. Geographically, a lot of efforts are underway to address it, and that’s probably why it’s time for consolidation.
“We think the next step in the market’s maturity is for there to be some kind of consolidation,” where the disparate fog efforts are consolidated to drive acceleration, Vasey said. “I think there’s a powerful combination [and] if we can get the industry to fully align on fog, I think we’re going to see a huge acceleration.”
Rhonda Dirvin, director of marketing in the Embedded and Automotive division of ARM, said the fog ecosystem is not in a position to open source any of its know-how just yet, but that’s a natural progression for the initiative in the long run.
Currently, Cisco, Foghorn, Nebbiolo and RTI are among the companies to offer fog-based solutions.
“I think 5G is even going to drive more need for fog computing,” she said in an interview. With 5G, there’s going to be quicker and more access to data and it would be useful to have more meaningful insight on that data, combined with Artificial Intelligence (AI) and machine learning.
It’s worth noting that the definition of fog computing is still a popular topic. Related to that is where the edge ends and the fog begins, but Dirvin cautions anyone talking about edge definitions to consider the source. A telecom engineer will have a different definition of edge compared to someone in the gateway community, for example.
Even those who were intimately involved in the early days of the OpenFog Consortium stay clear of making blanket statements about the edge and fog.
“Notice that there’s no clear boundary to say this is the edge and that’s not,” said Mung Chiang, dean of the College of Engineering at Purdue University, during his keynote Tuesday. “There’s no clear binary choice to say where you’re done with cloud and where you’re done with the fog. In fact, what is fog to us probably looks like cloud to the ants living on the ground. But what is amazing is that it’s prevailing. It shows up everywhere and it can be confusing.”
The OpenFog Consortium’s OpenFog Reference Architecture has been adopted as an official standard by IEEE. The standard, known as IEEE 1934, relies on the reference architecture as a universal technical framework that enables data-intensive requirements of the IoT, 5G and AI applications.
The OpenFog Reference Architecture is based on eight core technical principles, or pillars: security, scalability, openness, autonomy, RAS (reliability, availability and serviceability), agility, hierarchy and programmability.
Charles Byers, senior technical lead and platform architect at Cisco, clarified during a session Tuesday that the “edge” tends to refer to a single layer in the region of gateways, maybe at the base of a cell tower or at the base of a LoRa hub or home network. Fog touches multiple layers and there are advantages to having more layers than the one thin edge, he said.
Fog also tends to be multi-tenant and horizontal. That means that fog “doesn’t really care whether it’s smart highways, smart cities, smart grid, agriculture, mining, medical, education, retail,” or any other IoT vertical. Fog is supposed to be equally suitable to all of those verticals, Byers said.
The open fog initiative certainly got a vote of confidence from one popular computer nerd who was feature in a fireside chat Tuesday: Apple co-founder Steve Wozniak.
“I didn’t really know about fog,” Wozniak commented, noting that he supports any attempt to avoid vendor lock-in and give the consumer the most choice. “To me, the consumer is more important” than the producer, and the way to grow the IoT market is by having open standards.