The explosive growth of smart devices has put a strain on internet infrastructure, forcing cloud-based companies to adapt and find new ways to serve their customers. Edge computing—the practice of storing and processing data as close to the end user as feasible—is a rapidly growing industry that is barely keeping pace with demand. Just a few years ago, an estimated 10% of enterprise data was processed on the edge, outside of cloud-based data centers. By 2025, this number is expected to reach 75%. To get a better sense of edge computing’s future trajectory, we must first explore the biggest drivers of its growth.
These days even wristwatches and refrigerators have IP addresses, and the prevalence of smart devices is projected to quadruple by 2025. Consumer applications such as smartwatches are only part of the growth of these devices, and the real power of Internet of Things (IoT) devices can be seen in manufacturing, healthcare, retail, and the military.
For example, wifi-equipped sensors monitor factory equipment for warning signs of failure, deterioration, temperature fluctuations, increases in pressure, vibrations, and all manner of datum. That information is passed on to computers for processing, but if the factory’s demands are too urgent, even the speed of light is insufficient if the data is being processed on the other side of the continent. Edge computing has brought this crucial number crunching close to home, where worrisome fluctuations in factory conditions can be recognized and responded to in real time.
Also read: 5 Emerging Cloud Computing Trends for 2022
Energy on the Edge
Take for example the huge ramifications edge computing possesses for the oil industry, which typically operates in remote locations all over the globe and distributes its product via thousands of miles of pipelines. There simply isn’t enough manpower to provide on-the-ground monitoring of such infrastructure (though weekly airplane flyovers provide some inspection). The real work is being conducted by pressure and electrical conductivity sensors, which can distinguish if a pipeline has been punctured or if its protective anti-corrosive coating has suffered even a pinhole scratch.
Traditionally, this vast amount of long-range data would be gathered and transmitted to a single—and often quite distant—observational hub, where operators could remotely open and close valves to prevent environmental harm. Under the paradigm of edge computing, an anomaly in pressure can be quickly diagnosed, and an AI-driven response can take the appropriate actions, such as shutting valves near a potential leak.
It’s remarkably easy to understate just how much growth potential there is for edge computing in this industry. There are hundreds of thousands of miles worth of legacy infrastructure that can be modernized and many new construction projects that will implement the latest technologies to ensure safe product delivery for an industry that has seen its fair share of scrutiny.
The Fourth Industrial Revolution
For the past decade, German factories such as luxury car maker Audi have been pushing their world-renowned manufacturing to the next level with computerization. This strategy, dubbed “Industry 4.0” by the German government, has seen global uptake, with many industries adopting its four principles: create interconnected machinery, implement information transparency, empower machines to make decisions autonomously, and create systems that provide technical assistance to human operators. All of these principles require a factory to be fully loaded with IoT devices with low-latency demands.
In the case of Audi, a single plant can produce 1,000 vehicles each day, with more than 5,000 welds per car. That’s 5 million welds each day that require inspection—an infeasible number to manage. To solve this problem, Audi partnered with Intel to implement inline inspection of each weld, powered by sensors on every welding-gun, analyzing data such as electric voltage, health of electrodes, metal types, and configurations of each weld, and processing that data on the edge in real time. Audi now seeks to apply the same edge solutions to other processes such as riveting, gluing, and painting. Their success story has paved the way forward for manufacturers all over the world.
Also read: Edge Computing Emerges as Next Big Cybersecurity Challenge
Bringing Healthcare to the Edge
Medical sensors, electronic health records, and digital imaging systems are pushing volumes of data into the cloud, where high-resolution images rack up cloud storage fees by the gig and eat bandwidth along the way. Large hospitals running hundreds of thousands of sensors would be better served following the Audi model and keep that data close to home.
One healthcare provider partnered with Red Hat to create a real-time sepsis diagnostics solution, powered by edge computing. A traditional sepsis diagnosis eats precious time as charts undergo manual review and the condition worsens significantly by the hour. By creating a distributed network of medical sensors, diagnostic tools, and edge data processing, the provider has been able to detect sepsis nearly a day earlier than they had with traditional means, saving an estimated 8,000 lives to date as a result. That model, and others like it, stand to save many more lives as they are applied in innovative ways in hospitals across the country.
National Security on the Edge
Militaries operating in distant theaters may not have the luxury or even the desire to exchange data with the cloud. Edge networking, sourced from carriers or forward operating bases, is bringing enhanced data security and availability to soldiers in the fog of war. Changes in weather, breakdowns in machinery, and actionable intelligence can now be quickly assessed, relayed, and acted upon. The US military is also adopting edge computing at the homefront, responding to pandemic situations that pushed workers out of the office and decentralized their computing needs.
How Far Do We Push the Edge?
The future outlay is clear for the edge. Computers are running more complex calculations, sending more data, and often sending transient data that doesn’t require cloud storage. Devices are getting smarter, and there are more of them out there. Companies and consumers want to access their data quickly, reliably, and securely. Edge computing satisfies these growing demands, and it will grow accordingly.
Read next: Top Edge Computing Companies of 2021