Living On The Edge (Part II): What’s Driving Edge Computing? – Forbes

In this second article of a three-part series on edge computing, we will dig deeper into the concept of edge computing and explore what is driving the move to the edge as well as real-world examples.

What And Where Is The Edge?  

As I covered in part one, edge computing is a distributed, progressive computing paradigm that decentralizes data storage and processing, bringing it closer to the data source. By progressive, we mean that the computing happens closer to where the data is generated and not on the centralized cloud where the data is shipped over long network hops. The term “edge” covers a broad range of autonomous and interconnected devices and platforms where data is being generated and consumed outside of the data center (i.e., traffic signals, watches, smart meters, lightbulbs, smart buildings, insulin monitors, vehicles, etc.).

Because edge computing is an infrastructure architecture approach, it requires time, careful planning and resources to implement. With edge computing growing to an estimated market size of nearly $29 billion within the next five years, the time to implement edge computing projects is now.

What’s Driving Edge Computing? 

There are both business and technical drivers behind the growth in edge computing. On the business side, enterprises are continually looking for a competitive advantage through better customer engagement and lower cost of operations. From a technical perspective, it is the result of the rapid expansion in compute, storage (SSDs) and networking (5G, 100GigE) infrastructure that are part of mobile computing and the internet of things (IoT). The proliferation of mobile and IoT applications have highlighted the limitations inherent in centralized models. Edge computing is being used to address both of these areas.

The primary drivers behind edge computing are:

• Mobile and IoT applications. The exponential growth of mobile and IoT applications continues to fuel the need for highly available, low-latency, high-performance, secure, scalable platforms to process all of the data that’s being generated and consumed at the edge. Although the basic requirements are the same, over the past 10-20 years, mobile and IoT have significantly outgrown the traditional solutions that were being implemented to support them.

• Cloud limitations. Cloud is limited by its centralized processing model, which introduces cost, bandwidth, performance, security and regulatory issues. There is simply too much data being generated today and not enough bandwidth to allow it all to be transferred to the cloud for storage and processing. For real-time control systems like automated vehicles, factory automation and field inventory control, cloud network latency is too high, and reliability is too low to be practical (you don’t want your autonomous car sending data to the cloud in order to make real-time decisions). Finally, even if you could overcome data volume, network latency and reliability issues, attempting to infinitely scale a centralized compute-process-and-storage model is simply not economically feasible.

• New computing capabilities. As a result of the evolution of technology over the last 30 odd years, we now have far more compute and storage at the edge than ever before. For example: specialized processors that can handle local data processing for security (iPhone/Apple), artificial intelligence (Alexa/Google), video (Nvidia) and autonomous vehicles. Distributing compute, processing and storage tasks to these new, more capable devices can address many of the issues identified above, providing an improved, richer, more immersive customer and user interaction at the point of experience (PoE).

• Architecture as a competitive advantage. The rise of microservices and cloud deployments had started to turn application deployments into a somewhat commoditized activity, with little ability to differentiate from the competition (basically just spin up more cloud resources, as needed). Using edge computing as a competitive advantage, companies can extend the capabilities of their applications by leveraging topologies that can provide them with unique differentiation.

Examples Of Edge Computing Applications 

There are several successful edge computing use cases in live environments today:

• Healthcare: Mobile applications such as those created for concussion detection help bring healthcare to the sidelines and the field for athletes, directly connecting patients with their care teams and automated AI services.

• Supply chain: Just-in-time manufacturing is a thing of the past. Just-in-time supply chain enabled by virtualized custom design and manufacturing processes is allowing suppliers to offer options that their competitors just can’t match.

• Field workforce: People working in the field, be it sales or field service, need secure corporate data in often-disconnected environment. Edge computing enables 100% uptime, regardless of one’s internet connection.

• Autonomous vehicles: This includes automobiles, but it also extends to drones, submersibles and satellites. Here, AI processing and decision making is added to just the management of data.

• Travel and hospitality: Providing rich, personalized customer experiences aboard planes, trains and ships is the next phase of evolution.

• Manufacturing: Industrial IoT is a vast area where process automation is nearing the next levels of advancement as manufacturing equipment is more connected now.

• Video and gaming delivery: Video streaming services and gaming platforms leverage edge computing to accelerate and enhance content delivery by using storage, image processing and even predictive algorithms at the point of delivery.

Every industry has applications of edge computing that have both bottom-line advantages (cost savings, operational efficiency, etc.), as well as great potential for growth through differentiated offerings that give them that competitive “edge.”

What’s Next? 

Stay tuned for the final article in this series that will touch on the requirements of edge applications and how to get started turning edge computing into your competitive advantage.

This UrIoTNews article is syndicated fromGoogle News