September 3, 2020
Offshore drilling, gaming and medicine are all pivotal edge computing use cases, according to members of Protocol's Braintrust.
COO, Customer Operations at VMware
Autonomy and speed of decision making is the essence of edge computing. Edge and IoT captures data from the physical world instantaneously, while handling the captured data intelligently and interpreting it in real time. A real-world example would be offshore sites such as oil rigs. They're in the middle of the ocean, and they have to function alone. There's no data center nearby, there are no IT teams there, and the oil rig has to work all the time, be resilient, have the intelligence but also stay connected to the private and public clouds.
The data produced by edge and IoT helps enterprises improve business agility, create efficiencies and deliver better customer experiences. The edge is where compute acts on data in real time. Edge and IoT allow companies to analyze the physical world to be more innovative, successful and proactive for their customers.
CTO at Puppet
We are in a make or break moment for edge computing — the "hype" is only exacerbated by an increasingly remote workforce with more devices clogging enterprise networks than ever before, and the expectations of companies are also higher with respect to the impact technology can and will have on their organizations. A lot of innovation is happening around the world right now to tackle solutions for the edge, and the path to get there.
Edge computing is tightly coupled with 5G. As more telecommunication companies around the world start to roll out 5G networks, the possibilities become more interesting. More and more diverse workloads can start to move to the edge, powering where we go in new and innovative ways. Today, there are startup accelerators around 5G and, more importantly, work happening at the federal level to bring standardization to 5G networks, which in turn will have a direct impact on what can happen on the edge.
However, as more workloads move to the edge, security (and its sibling compliance) still remain a key concern for organizations around the world. Just as cloud migration has fundamentally changed how IT organizations think about application development and security, the transition to edge computing will also require a completely new way to build, deploy and maintain applications securely. What will make edge computing a success is an enterprise's ability to manage and secure all the endpoints — from the edge to the data center. Securing hundreds (if not thousands) of devices is complex, and organizations will need to adopt new technologies and strategies to truly secure their data on the edge in order to make this new trend a reality at scale.
Nancy Ranxing Li, Ph.D.
Director, Edge Cloud Platforms at Cox Communications
Edge computing will be the future of the "new Cloud." Many hyperscalers are making moves to extend their cloud offering to the edge. Imagine, if you can put the cloud in your backyard, what innovation can be empowered on the edge?
When developing applications on the edge, it's just like when you develop them on the cloud, but with lower latency, better customer experience and potentially lower total cost.
The most popular use cases on the edge are gaming applications. Low latency is extremely important for video game players and esports. There is also a huge financial incentive for game providers to offer their games on the edge. Gaming content providers usually pay high fees to offer their games through gaming consoles, such as Xbox. If they can offer the same game, such as Fortnite, on the edge, they can reduce their distribution cost significantly.
AI inferencing is also an early adopter of edge computing. Training AI models will still happen in the cloud, given that training the models will take a long time. However, inferencing can and should happen on the edge, if customers expect fast response and have a large amount of data to process, such as real-time processing images and videos.
The edge computing space is so exciting. I can't wait to see more innovation empowered by the edge.
GM, IoT Ecosystem at Intel
AI accelerated computer vision at the edge is transforming industries and enriching lives today.
Take for example one of retail's biggest challenges: inventory distortion (having too much or too little inventory); it's a trillion-dollar problem. Using autonomous robots and IoT-enabled data, retailers ensure the right product gets in the right hands, at the right time. In health care, edge computing enabled with AI can analyze medical images and provide insights to clinicians. In smart cities, we've deployed edge computing architecture to optimizing traffic flow, so you don't have to sit at an intersection any longer than needed. In industrial, edge computing is enabling manufacturers to detect defects before the product goes through the full flow, or quickly reconfigure their manufacturing line in response to changing consumer tastes.
The set of connected systems and devices from the edge to the cloud delivering intelligent insights will continue to enable amazing new technologies like autonomous cars and robotics, as well as build more predictive, intelligent and resilient communities.
Principal Technical Evangelist at F5 Office of the CTO
Real-life use cases for distributed compute (edge) tend to fall into two broad categories:
- Localized processing of data too large or time sensitive to efficiently transport and process centrally
- To mitigate connectivity issues
In the case of processing data, we can look at health care and the need to process MRI data. This data is large enough in size to incur significant costs and time to transport to a central location for processing. Using edge, such data can be processed more efficiently with relevant extract images transported to be stored centrally. This use case extends to manufacturing and natural resource firms in need of more immediate analysis and response to data generated by critical physical components. The latency incurred by relying on a centralized processing system can be hazardous to workers and to equipment.
The second general use case, while having less of an impact on safety, is nevertheless key to digital business. Retail locations often share networking which can have low availability or be subject to periodic buffering. Retail locations fall back to paper receipts when network connectivity fails, which is slow, expensive and very prone to fraud. Retail needs local advanced proxies for core transaction processing apps that are capable of securely storing transactions for up to a few hours' worth of network outage to be securely uploaded as soon as connectivity is restored.
In both use cases, edge mitigates the costs, time and risk associated with central processing of certain types of data.
Wendy M. Pfeiffer
CIO at Nutanix
Edge computing will thrive only if the platforms that enable it are flexible and performant enough to make use of the rich data constantly being generated on devices at the edge. The challenge is similar to the challenge we had at Yahoo 20 years ago: How do we collect and analyze data generated on client browsers in near-real time in order to enhance the human-machine interaction, while also influencing and prioritizing our longer-term product roadmap to incorporate these learnings into our ecosystem?
Ultimately, the platforms most adept at meeting this challenge will shape the ways in which we apply edge computing to business and society.
See who's who in Protocol's Braintrust (updated Sept. 2, 2020).
Questions, comments or suggestions? Email email@example.com
Kevin McAllister ( @k__mcallister) is a Research Editor at Protocol, leading the development of Braintrust. Prior to joining the team, he was a rankings data reporter at The Wall Street Journal, where he oversaw structured data projects for the Journal's strategy team.
More from Braintrust