Cloud + Edge: Three myths and a truth

The term ‘edge’ can be a bit cloudy. Most people think of edge in the form of self-driving cars, autonomous robots, 5G, or automated retail—innovative solutions that are revolutionizing the way we experience the world. And while edge is certainly responsible for powering those newer capabilities, edge is also present in your everyday life and is available at your fingertips.

Edge will drive the next wave of digital transformation so we all should know the opportunities and benefits it can provide our businesses. Here are three myths and a truth about edge computing to help bring clarity to what it is and isn’t.

Myth 1: Edge requires 5G

Edge can leverage 5G, but edge does not require 5G. 5G is one of many “on ramps” to a network, with others including Wi-Fi, fixed broadband, and more. Nonetheless, it’s the use case that determines which on ramp is best to use. For example, if a warehouse is using a lot of robotics and autonomous vehicles that need low latency to operate, a private 5G network may be the right fit. However, if a retail store needs cameras for loss prevention or technology for self-checkout, Wi-Fi or even a wired LAN may be most suitable. Both scenarios are examples of edge computing. The key point here is to not let the network drive the design. Define your use cases and find the best on ramps.?

So, do you need 5G to take advantage of the edge? Sometimes, depending on the use case.

Myth 2: Edge is only IoT (Internet of Things)

Many of the early use cases of edge computing revolved around IoT sensors. So, it’s understandable that there’s a perception they are one and the same. But there are many more ways to monetize edge than strictly IoT. IoT is just is one leg of the edge stool.

As specialized architectures, edge capabilities address technical limitations such as connectivity restrictions, high throughput of data, real-time requirements, local context-based decisions, no/limited human intervention, and content sensitive needs (regulation, security, or privacy). Edge use cases can be far more broad-based than sensor management. With edge:

• Retailers can use cameras and video machine learning, or beacons, or mobile apps to identify customers when they walk into the store, and then push out real-time offers to influence purchasing behaviors.
• Manufacturing companies can install high-resolution cameras that are trained to inspect products to ensure they meet standards. If an imperfection is identified, a warning is triggered for the product that needs to be re-inspected, not disrupting the flow of the other products.
• And companies can use machine learning to train cameras to recognize personal protective equipment. If an employee’s safety hat is missing, the camera will send data to the machine prompting it to turn off (in real-time), which will proactively protect the employee from injury.

So,?is edge just for IoT devices? No, it has many uses; some include IoT.?

Myth 3: Edge is only available on-premises

Actually, the opposite is true. Edge is an extension of cloud computing that is closer to users and devices—and most critically, data sources—that allows you access to real-time data to make decisions. It’s helpful to think of the cloud as a continuum of capabilities that span from public to edge—and everything in between. This cloud continuum includes different types of cloud ownership and location, from public to private or hybrid to co-location to telco to multi-cloud. The cloud is the “hub,” and the best place to develop applications, while the edge serves as the “spokes” that provide lower latency, faster data processing, and improved user experience.

So, is edge limited to on-premises? No, edge is part of the cloud continuum and can be physically located on-site or off, or in multiple sites.

Truth: Workload placement is critical

There is a fungible line from the cloud facility to the edge—which may include the network edge, radio edge on a wireless network, and further points along the way. It can be relatively easy to build applications for a big data center. But the farther the distance from the data center, the less resources and bandwidth there is to run applications, which means it requires more specialized resources to develop them. So, not everything belongs at the edge, and not everything belongs in the cloud—that’s why workload placement is a critical dimension of your cloud strategy. Scalability, storage, and network all need to be taken into consideration.

Decision-makers must think carefully about where applications need to reside along the continuum given the tradeoffs. In a retail environment, for example, what workloads are needed on a device a worker carries, versus in the store or centralized? Some retailers are beginning to redesign their in-store experiences from an edge perspective. In the future, one could imagine retail stores with smart shelves that swap out products based on the time of day to ensure that more popular items are being seen by a certain customer base when they visit.

Simply put, if the use case has specific requirements that edge can fulfill, it’s a good fit. This may include a need for low latency to process data incredibly fast, or, if a mission critical application must be ‘always on’ and processing independent of connectivity. It’s great for optimizing digital consumption such as incredibly fast video caching or extended reality experiences. And it makes sense when information needs to remain local due to compliance requirements or to protect privacy in the processing of sensitive data.?

Every company should be asking itself when, where and why it’s important to have edge as part of its cloud strategy. At Accenture, we’re investing in research, experimentation, and co-creation to understand and demonstrate the value edge can provide organizations. We work with our clients to help define the appropriate edge applications and then architect them effectively to take full advantage of the benefits the edge provides. Please reach out to me to discuss how you can maximize the value your company achieves from edge throughout its cloud journey.

For more information on bringing edge into your cloud strategy, read Accenture’s The Cloud Continuum report, and the three tech enablers of the Cloud Continuum. Learn more about how edge is being used in the public sector here.