Edge Computing is an increasingly important concept in today’s world of intelligent distributed computing. As tempting as it is to dismiss it as just another trendy buzzword, it is at the heart of technology powering high-growth sectors such as Internet of Things, Cybersecurity, Autonomous Driving and voice-driven applications. In this post, we dive into what it means, how it works, and how it fits into broader technology and business strategies.
What it is
Unlike the earlier client-server computing model, the premise (and promise) of cloud computing was to consolidate all intelligence into a centralized cloud-based computing model. This provided the benefits of core algorithms being centralized and driving all application and business logic, relying on client applications only for superficial user interfaces. This was further enhanced by Artificial Intelligence (AI) models which also benefitted from centralized data assets.
But in today’s highly connected and real-time landscape, many decisions need to be made quickly and without the latency of data transmission to a centralized engine. This type of “delegation” allows devices at the “edge” such as sensors, cameras or input devices, to make certain decisions quickly and autonomously. In a world of billions of connected devices, this approach greatly minimizes network traffic and mitigates the risk of network problems or latencies resulting in application failures.
Currently, around 10% of enterprise-generated data is created and processed outside of a traditional centralized data center or cloud. By 2022, Gartner predicts this figure will reach 50%.
Edge computing is imperative for systems that require higher levels of immediacy. For example, in Security Operations, edge-based intrusion detection systems need to detect and react immediately based on observed patterns in network traffic. In home security, cameras can detect moving objects and track them autonomously. Manufacturing systems are also perfect examples of edge-computing at play with individual subsystems and sensors analyzing, detecting and reacting of their own volition.
Perhaps the perfect example, and also one that is easy to understand, is autonomous cars, where cameras, sensors and other intelligent elements need to perform highly intensive processing around contained sets of data – such as distinguishing a pedestrian from a lamp-post – and make rapid life or death decisions – without the need for centralized processing or massive data transfer. Indeed, autonomous cars would not be possible without edge computing.
Edge computing also introduces its own set of challenges. With more elements operating autonomously, security becomes even more crucial and approaches need to be multi-faceted, detecting not just network attacks but also localized code-corruption, version management, and firmware vulnerabilities. Innovative data strategies are needed to ensure that localized algorithms benefit from real-time information, and to ensure that disparate intelligent subsystems are all working in concert.
If you conduct business in any of these areas, or if your technology stack involves a distributed network of devices, sensors, cameras etc., Edge Computing should be a critical part of your digital strategy in 2018-2019. If you’re interested in taking an even deeper dive into the methodologies behind edge computing, or are considering edge computing for your business, reach out and our team would be happy to connect.