The shift to digital is creating a hyperconnected world of devices, sensors, phones, networks and connected infrastructure. The number of devices and the amount of data that will be collected and processed is expected to grow exponentially. Technicians must rethink how their applications and systems evolve to deliver powerful, scalable, and performant experiences.
Let’s discuss the actual opportunity for the business, and the components needed for an edge computing technology strategy to support it.
01
What business opportunities does edge computing create?
Business use cases for edge computing are evolving rapidly. Here are a few current trends.
Immersive retail experience: more responsive and intelligent point-of-sale terminals; connected shopping kiosks; immersive product reviews; mobile inventory inquiry and product inquiry applications.
Industrial Measurement: Industrial facilities connect their machines to track usage, provide maintenance updates, and notify employees of problems.
Supply Chain Inventory Management: Distribution centers use cameras with artificial intelligence to analyze inventory levels and keep them on track.
Medical Monitoring: Healthcare facilities track patient information to improve diagnosis, prescribed doses, and changes in patient condition to improve treatment.
02
What are the key technical architectural considerations that support this strategy?
Equipment Precautions
End-user or system devices may have compute, storage, and networking capabilities, or they may be lightweight and lack these capabilities. These devices include point-of-sale terminals, sensors, tablets, mobile phones, cameras, vehicles, appliances, and more.
Due to availability requirements, devices must submit data that cannot be processed to an edge location within a metropolitan area or an alternate metropolitan area to meet availability requirements. Application providers must design device routing, monitoring and management, and over-the-air update capabilities.
Edge Computing Layer Considerations
Edge computing is a distributed computing design that brings computing and data storage closer to the edge of the technology provider's network, resulting in a more responsive experience on the device. To provide a more responsive experience to end users or systems, application designers are deploying edge computing solutions in metropolitan areas closer to devices. Key benefits of edge solutions include low latency, high bandwidth, data cleansing and AI processing, responsive analytics, and trusted computing and storage.
Applications should expose APIs for edge and edge provider applications for use by third-party and top-level applications. This design will simplify the interaction between application developers and the telecommunications network between the application and the device itself. Exposing information such as the user's device and device location can improve use cases.
Developer tools can be used to more easily create APIs between devices and edge solutions, as well as technology provider environments. Application providers can use content delivery network technology to distribute APIs and other resources. This strategy provides robust, rapid deployment speed to all edge locations in metropolitan areas. It also makes the solution easy to install and manage.
Some use cases require the deployment of various applications at different sites within and across metropolitan areas. Consider a variety of edge sites deployed throughout a metropolitan area that communicate with each other to provide optimal coverage similar to a mesh network in the home. In this case, a distributed cloud is valuable as an execution environment for applications across multiple locations, including connections managed as a single solution.
Network considerations
The most important network considerations are high-speed, high-bandwidth, and reliable network connectivity from connected devices to edge locations, and from edge locations to cloud data centers. Customers will reject disconnections, cached and pixelated videos, and frequent disconnections. A responsive and sometimes immersive experience on the device is a key factor in customer satisfaction.
Networking capabilities on the device must route data to the nearest edge location where the application resides and help meet the application's needs, providing a better customer experience. This process should be a simplified mechanism, such as distributed anchors or multiple sessions, and consistent with a standardized approach. Edge locations within metro areas can also be networked for greater connectivity and responsiveness, much like a mesh router system in a home. These locations can also provide cross-city connectivity for failover and disaster recovery needs.
Data center considerations
Data center providers are creating configuration options to host edge computing systems based on the needs of the application. These data centers can host entire applications, or they can be used in a distributed model between devices, edge, and hyperscaled cloud. If the application requires extremely low latency, all processing may be located in a data center within the same metropolitan area as the device. For example, point-of-sale terminals can benefit from this configuration due to latency requirements.
Depending on the processing and latency needs of the application, edge locations can be met in various configurations. Smaller data centers can also be deployed throughout metropolitan areas to provide redundancy, lower latency, a responsive experience to the location of the equipment, and simplify space and power considerations. These edge data center configurations include micro data centers (less than 5 racks), mid-sized data centers (10-120 racks) and hyperscale data centers. These options have the required physical protection, cooling, networking, and power capabilities.
Applications can distribute the heaviest algorithm and storage processing within centralized cloud data centers located outside of metropolitan areas at device and edge processing locations. This design strategy expands processing capabilities and can also be used to achieve regulatory and privacy compliance, as data processing and storage can be placed in areas that meet the requirements.
03
in conclusion
Supply chain challenges are disappearing, driving the rapid growth of hyper-connected devices and new application systems. The mentioned technical design principles and advancements create new business opportunities that would have been considered unrealistic just a few years ago. Understanding how this distributed computing model emerges to create a better experience for customers and employees is critical.
