Edge computing is here to stay – here’s why
Think of the several CCTV cameras located on the premises of your office building through the course of a night. Several hours’ worth of video data is recorded and then sent to connected Cloud servers to be processed – revealing very little data worth processing. This begs the question: does anyone really need the several hours of footage showing just empty corridors and rooms? Now instead of this, imagine an office – an IoT-enabled environment – where the data from an AI-enabled camera is processed by an engine in real-time to send only relevant data (unusual activity, for example). This is what Edge computing provides – data being processed at the ‘Edge’ – i.e. the source.
Edging over Cloud
As part of the digital transformation process, Cloud computing is already set to see a significant boost owing to several business resources being supported remotely on the Cloud. However, it is not just the Cloud that would see this boost. Edge computing, too, could witness similar boost in certain environments where data processing can be carried out at the point of collection, thereby saving time, bandwidth and costs of massive data storage and processing. Essentially, Edge computing is an expansion of the current Cloud computing architecture – ‘an optimized solution for decentralized infrastructure’.
According to research from Gartner, by 2022, almost half of the total data generated by enterprises will be processed beyond centralised Cloud data centres, within closed IoT environments via Edge computing. By 2025, the total number of IoT-enabled devices is expected to reach the 75-billion mark. In fact, research from Analysys Mason even suggests that enterprises may spend up to 30% of their total IT budgets in setting up Edge computing environments within the next three years.
Edge computing is essentially a distributed IT network architecture enabling mobile computation of local data. By doing so, it decentralises processing from dedicated Cloud centres, thereby opening up bandwidth as well as reducing latency and storage requirements. According to media reports: “The concept dates back to the 1990s when Akamai (headquartered at Massachusetts) solved the challenge of Web traffic congestion by introducing Content Delivery Network (CDN) solutions. The technology involved network nodes storing static cached media information at locations closer to end-users.” Firms across verticals can expect anywhere between 10-30% cost reduction upon adopting Edge technologies, additionally saving operational costs by about 20%.
Industries on Edge
Here are some of the ways that industry has implemented Edge computing:
Hazardous work environments: For firms in the construction, manufacturing, mining, and oil and gas sector, Edge computing can assist greatly in improving worker safety and asset management. Working on solely a Cloud-enabled environment could potentially halt the production process or even prove to be rather dangerous (for workers in hazardous environments) in cases of power outages or network disruptions.
Hence, an IoT-based environment where local devices can interact and share relevant updates and notifications with each other covers all bases considerably. Using low-latency Edge computing thereby ensures worker safety, increased productivity and better management of projects. Additionally, it can also be used to monitor environmental effects and pollution in real-time to make sure safety thresholds are not breached. Thus, being on the Edge can literally save lives!
Automobiles: A classic example of the use case of Edge computing is in electric vehicles. This includes monitoring battery (and engine) health and predictive maintenance, optimising charging stations and processes, smarter traffic management on the roads, personalised in-car entertainment and improved vehicle security (via tools like multi-level authentication).
However, arguably the greatest automobile evolution of the century comes from the advent of self-driving (autonomous) cars. By some estimates, these vehicles generate as much as 4 terabytes of data per hour. Most of it can be processed within the car itself and does not need to be transmitted to the Cloud. Edge computing is essentially the “co-pilot to onboard computers in driverless cars”. It is designed to onboard massive volumes of data from car sensors and cameras and process them instantaneously, making decisions at the ‘Edge’.
Logistics: In some passenger and freight railroad operations, Edge computing can help create a smart IoT platform to digitise old and ailing infrastructure. An IoT-Edge combo is being deployed at several railroad crossings that could reduce associated maintenance costs. The volume of data being transmitted about aspects like the health of gates, signals, bells or batteries at crossings can be reduced to simple one-line error messages that can be sent over low-cost LoRa (long-range radio) networks. This eliminates the need for expensive data plans and network costs, especially if the data had to be transported from a device to a network. Using simple ‘if-then’ logic, or even continuously-learning AI algorithms which understand the issues and build short summary reports, can greatly increase efficiency, while simultaneously reducing even medium-run costs.
Retail: Currently, retailers use beacon technologies to collect customer information when they walk into stores. Beacons are essentially small wireless transmitters that use Bluetooth technology to transmit information to smart devices nearby, making actions like location-based searching much easier.
Globally, businesses and marketers spend billions every year on online marketing. Using beacon technology, businesses get a better idea of tracking how well their digital ads (given a user’s online activity relating to Google Ads, for example) are doing to drive customers physically to their stores. Edge computing goes above and beyond this by sending shoppers targeted promotions and sales items as they walk in the doors. It also allows for greater operational activities for businesses, enhancing both real-time decision making and improved autonomy in localised operations.