I recently attended my son’s graduation. I was one of scores of parents trying to capture and share the moment via my smart phone. My wife strategically placed in one area of the arena and I in another, we texted each other to report my son’s location, trying to document the key moments leading to diploma-in-hand.

It didn’t happen, despite our best efforts. With so many people crammed into one location vying for connectivity and bandwidth, our communication was delayed by roughly 20 minutes. We missed “the” major moment and the joy of sharing, due to latency issues.

Tomorrow’s scenario: A better way

It didn’t have to happen this way. Smart ad-hoc networking could have utilized connectivity and distributed “compute” power amongst devices in the crowd of attendees. Using edge computing to offload low-latency, high-bandwidth tasks to a device cloud, would have eliminated the lag time that frustrates so many of us in remote vacation areas or crowded urban events. It’s a sharing economy model, in which—with permission—devices in a given area that have spare capacity can function as server and client interchangeably.

For communication service providers (CSPs), retailers and other entrepreneurially-minded entities, this type of edge-computing scenario opens new service opportunities—enabling them not only to serve customers better but to monetize data. In the same way ride-sharing services utilize spare commute capacity, or numerous online marketplaces capitalize on spare lodging capacity, savvy companies can bank excess processing power. The processing power in smartphones has doubled year over year for some time now, creating spare compute capacity that potentially can be pooled. This extended Mobile Edge Computing (MEC) provides ad-hoc, device-based cloud networking that could be a cooperative network between the devices being orchestrated or through a master orchestrator.

Regardless of the set-up, bringing smart phones into the sharing economy is a win/win. Consumers eliminate lag times and signal issues, while communication service providers and other businesses solve a customer problem, engendering loyalty and supporting profits. Yet today, barely half (52 percent) of executives Accenture surveyed see more people-centric experiences as a benefit of edge computing.i

The flip side is that traditional edge computing comes at an increased cost. It needs to be managed—and potentially augmented with creative architecture solutions to be truly viable.

Beyond venues, think of the possibilities for emergency relief in disaster recovery situations. When Hurricane Harvey hit Houston in 2017, a device cloud could have allowed citizens and emergency workers to share resources, helping direct those in imminent danger to safety—despite network issues due to the storm.

Edge computing on smart phones opens a world of possibilities.

Four key principles make edge computing work

To take advantage of this concept, business strategists and technologists must come together to create workable solutions. Four key areas should guide design:

Is it user friendly? At my son’s graduation, I needed an intuitive app so I could invite friends and family to network at the ceremony. We needed to pool smartphone capacity with quick setup and easy disbanding. Good design would have looked like this: I click on an app. It asks me who I need to connect to, showing me all nearby available parties who want to share. I get to see how much available battery each has, what their compute capacity is and how long they will be near me/willing to share. And of course, I can see how trusted they are based on security levels. I choose based on the information available and within seconds, the possibility of any lag time disappears.

Is it sharing-economy ready? Eighty-five percent of executives agree enterprises must balance cloud and edge computing to maximize technology infrastructure agility and enable intelligence everywhere.ii Hardware architectures matter in this scenario—invisible to the user but essential to success. They need to support renting out partitions of processers and compute resources that can be transferred and locked into a pooling mode for a specified period or task.

Is it orchestrated? At graduation, maybe Grandma could not attend but wants to feel a part of the ceremony. Users should be able to create an album or video in real-time to send to her. This requires dynamic, intelligent task allocation through an orchestrator. This scenario should take advantage of existing scheduling algorithms or variants, leveraging and re-architecting for dynamic ad-hoc networking.

Does it build in AI-based, proactive security? Sharing in a device cloud requires airtight security. With the proliferation of cloud services and with more IoT device classes coming online, the need to design bottom-up security is increasingly important. Comprehensive security can be effectively addressed by augmenting it with machine learning and AI-based techniques that can identify threats, adapting to address any emerging vulnerabilities.

Edge computing on smart phones opens a world of possibilities. Wise CSPs and partners are collaborating already to benefit from the convergence of emerging technologies, creating yet another set of monetizable use cases to drive innovation for consumers. Acting now ensures your company is a part of the ecosystems creating this innovation, rather than an outsider looking in on a missed opportunity.

i Accenture 2018 Tech Vision research

ii Ibid

Muzaffar Khurram

Senior Manager – Accenture Strategy, Communications, Media and Technology

MORE ON THIS TOPIC


Subscription Center
Stay in the Know with Our Newsletter Stay in the Know with Our Newsletter