As the cliché goes, life on the edge is for well-balanced people.
A lot of the buzz at GE Minds+Machines 2017 in San Francisco last week was about edge computing. GE unveiled Predix Edge Manager for GE’s prized Predix software platform, which lets customers support fleets of edge devices – up to 200,000 connected devices.
Be forewarned: Despite the hype, edge computing is still in its adolescent stage in need of companies culturally committed to digital transformation as well as hard-to-find technical skills, either via employees or digital-savvy partners.
“There’s a desire and there’s a capability, and those two aren’t married yet,” says Rob McKeel, CMO at GE Power.
Edge computing is where data collection, analysis and actions happen on the machine – that is, the machine doesn’t have to loop back to the cloud or datacenter to process the data. While edge computing solves latency and cost issues associated with centralized processing, its greater value is in the form of clear business outcomes.
Edge computing essentially crunches the time gap between data, insights and actions into mere milliseconds. In a blog post on Machine Design, McKeel cites a windfarm’s use of edge computing. When the wind changes direction, he says, edge software can collect and analyze this information on the fly and prompt the entire field of wind turbines to adjust the pitch of their blades and capture more kinetic energy.
Related: Zero One: Rise of the IoT Architect
Forrester calls edge computing one of the top technology trends to watch over the next three years:
- In 2018, companies will struggle reconciling edge computing with operational and enterprise technology.
- In 2019, companies will pilot edge computing use cases yet run into security and regulatory challenges.
- In 2020, leaders will emerge and have the edge on serving customers. (Yes, pun intended.)
“The expansion of computing from data centers and cloud to the edge is still in the dawning phase, as many technical and cultural obstacles stand in the way of success,” say Forrester analysts Brian Hopkins, Bobby Cameron, Ted Schadler and Rusty Warner in a research note.
Specifically, companies venturing to edge computing will need skills in edge servers, cloudlets, intelligent gateways, fog fabric nodes, hyper-local stream data management, distributed data and analytics management architectures, Internet of Things analytics, etc.
In an edge computing environment, CIOs will become more like consultants, Forrester says. The CIO will need to be knowledgeable about edge computing hardware vendors, security and integration, cloud vendors extending their offerings to the edge, and, of course, strengths and weaknesses of service providers claiming to be edge experts. In fact, Forrester believes edge computing will spawn a new infrastructure market.
On the upside, edge computing might solve some of the security problems plaguing the Internet of Things.
Consider one of the benefits of edge computing: autonomous operation. Through edge technology, a distribution substation for solar power nodes has its own power supply and computing power and can actually operate like an autonomous utility, says Steven Martin, chief digital officer at GE Power.
“If you think about being able to segment at that level, it adds to the collective security of the network, because you have some level of autonomous operation pushed out to the edge rather than purely centralized control,” Martin says.
As for early days of edge computing, Martin argues it’s not early – at least not in the utility industry. GE Power’s utility customers are already deploying assets, such as natural gas turbines and transformers, with edge computing in mind. That’s because utility companies today have to look out to the farthest edges of what’s possible.
“These assets are going to live for 30 or 40 years,” Martin says. “It’s an asset that survives an entire generation. And so we should expect that the digitization of these assets is going to be a long pursuit.”
Tom Kaneshige writes the Zero One blog covering digital transformation, AI, marketing tech and the Internet of Things for line-of-business executives. He is based in Silicon Valley. You can reach him at [email protected].