Even with the cloud’s fantastic success in such a short time, another transformation could spur a faster, more dramatic change: edge computing. While it isn’t going to replace the cloud, it will reshape it in new ways. Said Ouissal, CEO and Founder at ZEDEDA, shares three recent innovations that strengthen the case for the edge.
When Amazon launched Amazon Web Services back in 2006, it kicked off one of the most transformational periods in modern computing. Over the past 15 years, cloud computing has become the thread connecting business technologies and driving innovation. The latest forecast from Gartner predicted a 23.1% increase in spending on cloud servicesOpens a new window and a global market of $332.3 billion.
Companies have been collecting data at the edge of the network for years, even before the rise of cloud computing. The difference brought on by the IoT trend over the past five years is that the data is no longer just collected – it’s now connected.
The sheer amount of data coming from the edge will force a change to the architecture of the current cloud, as billions of devices collect data from the world around them then connect to a network. At its core, the cloud’s benefits, agility, and capabilities now need to be extended to the edge of the network. Reasons for moving to the edge include lower latency, reduced bandwidth costs, and better autonomy, security and privacy. The latest staggering predictions are that 75% of all dataOpens a new window will be processed at the edge of the network by 2022, and by 2025 nearly one-third of all the workloads in the world will run at the edge.
See More: Life on the Edge: Solving the Optimization Problem
Three Innovations Rooting for Edge
Companies need to invest in flexible infrastructure combined with differentiated applications and domain knowledge to maximize edge computing opportunities. Here are three examples of how industries are driving the edge revolution.
1. Autonomous Drilling
Despite its popular image of roughnecks working on oil wells, the oil and gas industry is heavily data-driven and has been for years. There’s a tremendous amount of processing, computing and data analytics that happens at the edge.
The next step that edge computing can enable is autonomous drilling, where artificial intelligence at the edge can manage critical functions at the well without the need for human intervention. The current version is called an open-loop system, where sensors and software tell the drillers what’s happening, giving the team a chance to adjust on the fly to what lies ahead. The goal is a closed-loop system, meaning recommendations are automated, and machines take appropriate action. Meanwhile, in specific critical applications, these AI-driven insights will augment human activity.
Schlumberger’s DrillOps is one of the most advanced solutions for autonomous drilling. DrillOps Automate can autonomously drill a stand in a closed-loop system with embedded IoT and AI technologies, and DrillOps Advise can provide the driller with smart alarms and real-time advice. The result is improved operational performance at a lower cost while ensuring worker safety.
2. Self-Driving Car Maintenance
Like oil and gas, the auto industry is also set in the physical world, but with a growing number of connections to the cloud. The automotive service sector, in particular, is changing quickly to adapt to electric vehicles.
While most of the focus in this space is inside the cars themselves, we also see innovations at service centers. As vehicles become electrified and smarter, garages will move from oil changes to managing software and data-driven car maintenance. Regular and predictive maintenance will become critical, with local infrastructure hosting software updates for cars to meet data sovereignty requirements.
Connected vehicles create a massive amount of data every second, and it simply isn’t feasible to backhaul all of this data directly to the cloud. Local infrastructure at service centers can take advantage of aggregated data to evaluate battery performance and the overall health of vehicles. A subset of this data can then be relayed to the cloud to understand trends across an entire fleet further. Â
It’s also essential to distinguish between latency-critical and latency-sensitive workloads for use cases for connected vehicles. Autonomous cars are effectively data centers on wheels, and any latency-critical decisions (e.g., deploy the airbag or brakes) will always be made locally. Â
Meanwhile, latency-sensitive workloads will be delivered by a mix of in-vehicle systems, edge computing resources in surrounding public infrastructure and the cloud. These workloads include infotainment, mapping, augmented reality and C-V2X (cellular to vehicle) communication between cars and local infrastructure to supplement decisions made within the automobile itself. An example is to warn two cars about a potential accident or negotiate a safe crossing if both are approaching an intersection at a high rate of speed.
3. Industrial Automation
The manufacturing industry has already made great strides in taking advantage of edge computing capabilities. Industrial robots collect data as they work and make tiny adjustments in real-time to optimize operations. Private 5G is also enabling new use cases in the manufacturing sector, which drives even more of a need for edge computing. While 5G enables a high-bandwidth, low-latency local connection, you need to preprocess data on-site because you still have the same bottleneck between the facility and the cloud.
Telstra is an example of a company that has combined a private 5G network with edge AI workloads to maximize efficiency, maintain quality and keep workers safe. Such solutions ensure autonomous operations regardless of cloud connectivity, minimize decision-making latency, and reduce the overall amount of data that needs to be backhauled to the cloud.
Edge computing also enables manufacturers to drive more efficiencies in monitoring, managing, maintaining, and optimizing their factory assets. BOBST, one of the leading suppliers of machinery and services to the packaging industry, set out to automate as much as possible the process to maintain the OS versions and apply security patches to edge devices in the field. The result was increased quality and yield and reduced risks and costs.
At the Edge of a New Beginning
This is not the end for cloud computing, even if it sometimes seems to be. We’re still in the early days of edge computing, but the possibilities are eye-opening. The cloud and the edge are complimentary. It’s not a zero-sum game. But the edge holds the key to tomorrow’s hyper-connected world. In many ways, the edge is the next cloud to build.
What are the challenges that can impede the growth of edge computing? Share with us on LinkedInOpens a new window , TwitterOpens a new window , or FacebookOpens a new window . We’d love to know what you think!