What is edge computing?
Edge computing allows IoT devices to process and act on data in real or near-real time by processing data at the edge of the network
Edge computing explained
Edge computing allows devices in remote locations to process data at the "edge" of the network, either by the device or a local server. And when data needs to be processed in the central datacenter, only the most important data is transmitted, thereby minimizing latency.
Why do businesses use edge computing?
Businesses use edge computing to improve the response times of their remote devices and to get richer, more timely insights from device data. Edge computing makes real-time computing possible in locations where it would not normally be feasible and reduces bottlenecking on the networks and datacenters that support edge devices.
Without edge computing, the massive volume of data generated by edge devices would overwhelm most of today's business networks, hampering all operations on an affected network. IT costs could skyrocket. Dissatisfied customers might take their business elsewhere. Valuable machinery could be damaged or simply be less productive. But most importantly, workers' safety could be compromised in industries that rely on intelligent sensors to keep them safe.
How does edge computing work?
To make real-time functionality possible for smart apps and IoT sensors, edge computing solves three interrelated challenges:
- Connecting a device to a network from a remote location.
- Slow data processing due to network or computing limitations.
- Edge devices causing network bandwidth issues.
Advancements in networking technologies, like 5G wireless, have made it possible to solve these challenges on a global, commercial scale. 5G networks can handle vast amounts of data—going to and from devices and datacenters—in near-real time. (There's even a wireless network that uses cryptocurrency to encourage users to extend coverage to harder-to-reach areas.)
But advances in wireless technology are only part of the solution for making edge computing work at scale. Being selective about which data to include and exclude in data streams over networks is also critical to reducing latency and delivering real-time results. For example:
A security camera in a remote warehouse uses AI to identify suspicious activity and only sends that specific data to the main datacenter for immediate processing. So, rather than the camera burdening the network 24 hours per day by constantly transmitting all of its footage, it only sends relevant video clips. This frees up the company's network bandwidth and compute processing resources for other uses.
More use cases made possible by edge computing:
- A retail store 1,000 miles from the company's primary datacenter uses wireless point-of-sale devices to instantly process payments.
- An oil rig in the middle of the ocean uses IoT sensors and AI to quickly detect equipment malfunctions before they worsen.
- An irrigation system in a remote farm field adjusts the amount of water it uses in real time by detecting soil moisture levels.
Why is edge computing important?
From workplace safety to security and productivity, the benefits of edge computing are vast:
More efficient operations. Edge computing helps enterprises optimize their day-to-day operations by rapidly processing large volumes of data at or near the local sites where that data is collected. This is more efficient than sending all of the collected data to a centralized cloud or a primary datacenter several time zones away, which would cause excessive network delays and performance issues.
Faster response times. Bypassing centralized cloud and datacenter locations allows companies to process data more quickly and reliably, in real time or close to it. Consider the data latency, network bottlenecks, and diminished data quality that could arise when trying to send information from thousands of sensors, cameras, or other smart devices to a central office all at once. Instead, edge computing enables devices at or near a network's edge to instantly alert key personnel and equipment to mechanical failures, security threats, and other critical incidents so that swift action can be taken.
Greater employee productivity. Edge computing enables businesses to more quickly deliver the data that workers need to complete their job duties as efficiently as possible. And in smart workplaces that take advantage of automation and predictive maintenance, edge computing keeps the equipment that workers need running smoothly, without interruptions or easily preventable mistakes.
Improved workplace safety. In work environments where faulty equipment or changes to working conditions can cause injuries or worse, IoT sensors and edge computing can help keep people safe. For example, on offshore oil rigs, oil pipelines, and other remote industrial use cases, predictive maintenance and real-time data analyzed at or close to the equipment site can help increase the safety of workers and minimize environmental impacts.
Functionality in far-flung locations. Edge computing makes it easier to utilize data collected at remote sites where internet connectivity is intermittent or network bandwidth is limited—for example, aboard a fishing vessel in the Bering Sea or at a vineyard in the Italian countryside. Operational data like water or soil quality can be constantly monitored by sensors and acted upon when needed. Once internet connectivity becomes available, the relevant data can be transmitted to a central datacenter for processing and analysis.
Heightened security. For enterprises, the security risk of adding thousands of internet-connected sensors and devices to their network is a real concern. Edge computing helps to mitigate this risk by allowing enterprises to process data locally and store it offline. This decreases the data transmitted over the network and helps enterprises be less vulnerable to security threats.
Data sovereignty. When gathering, processing, storing, and otherwise using customer data, organizations must adhere to the data privacy regulations of the country or region where that data is collected or stored—for instance, the European Union's General Data Protection Regulation (GDPR). Moving data to the cloud or to a primary datacenter across national borders can make adhering to data sovereignty regulations difficult, but with edge computing, businesses can ensure that they're honoring local data sovereignty guidelines by processing and storing data near where it was collected.
Reduced IT costs. With edge computing, businesses can optimize their IT expenses by processing data locally rather than in the cloud. Besides minimizing companies' cloud processing and storage costs, edge computing decreases transmission costs by weeding out unnecessary data at or near the location where it's collected.
Edge computing hardware and networking
In edge computing, much of the processing power is physically located at or near where the data is gathered. Edge computing hardware often consists of these physical components:
Edge devices include smart cameras, thermometers, robots, drones, vibration sensors, and other IoT devices. Although some devices have built-in compute, memory, and storage capabilities, not all do.
Processors are the CPUs, GPUs, and associated memory that power edge computing systems. For example, the more CPU power an edge computing system has, the faster it can perform tasks and the more workloads it can support.
Cluster/servers are groups of servers that process data at an edge location, such as on a factory floor or at a commercial fishery. Edge cluster/servers are often tasked with running enterprise apps, enterprise workloads, and an organization's shared services.
Gateways are edge cluster/servers that perform essential network functions like enabling wireless connectivity, providing firewall protection, and processing and transmitting edge device data.
Routers are edge devices that connect networks. For example, a router at the edge may be used to connect an enterprise's LANs with a WAN or the internet.
Switches, which are also referred to as access nodes, connect several devices in order to create a network.
Nodes is a catch-all term used to describe the edge devices, servers, and gateways that enable edge computing.
What are some of the characteristics of edge hardware?
Edge hardware needs to be durable and dependable. Often, this equipment must be able to withstand extreme weather, environmental, and mechanical conditions. In particular, it often must be:
Fanless and ventless. With reliability being key, especially in industries where equipment malfunctions can halt production and endanger workers, edge hardware must be closed off from dust, dirt, moisture, and other matter that could compromise it.
Temperature resistant. Edge hardware is often placed outside in freezing, sweltering, and wet climates. Sometimes it's even placed underwater. Being able to withstand sub-zero and near-boiling temperatures is a must in many cases.
Impervious to sudden movements. The hardware needs to be able to withstand vibrations and shocks by machinery or the natural elements. Building these components without fans, cables, and other internal parts that can easily get shaken loose or break is essential.
Small form factor. With edge computers, compact is the name of the game. They often need to fit into cramped spots. Examples include smart cameras placed on walls, shelves, and ceilings and smart thermometers packed in shipping boxes.
Equipped with ample storage. Edge computers that collect vast amounts of data from edge devices can require significant data storage. They must also be able to rapidly access and transfer large quantities of data.
Compatible with new and legacy equipment. Edge computers, particularly those operating in production or factory settings, typically feature a variety of I/O ports, including USB, COM, Ethernet, and general purpose ports. This enables them to connect with both new and legacy production equipment, machinery, devices, sensors, and alarms.
Built with multiple connectivity options. Edge computers typically support both wireless and wired connectivity. That way, if connecting to the internet wirelessly is not an option at a remote commercial site like a farm or a ship at sea, the computer can still connect to the internet to transmit data.
Able to support several types of power inputs. Edge computers often support a variety of power inputs to accommodate the wide range of power inputs they may encounter in remote locations. They also require surge, overvoltage, and power protection features to help prevent electrical damage.
Protected from cyberattacks. Edge devices, which often cannot be managed by network administrators as rigorously as their on-premises and cloud counterparts, tend to be more vulnerable to bad actors. To help safeguard them from malware and other cyberattacks, edge devices must be equipped with security tools like firewalls and network-based intrusion detection systems.
Tamper resistant. Because edge computing devices are often used in far-flung locations where they cannot be consistently monitored, they must be built to be kept secure from theft, vandalism, and unauthorized physical access.
Cloud vs. edge vs. fog computing
Edge and fog computing are intermediary computing technologies that help move the data collected by IoT devices at remote locations to a company's cloud. Let's explore how edge computing differs from fog computing and cloud computing, and how the three work together:
Cloud computing enables companies to store, process, and otherwise work with their data on remote servers that are hosted over the internet. Commercial cloud computing providers like Microsoft Azure offer digital computing platforms and collections of services that companies can use to reduce or eliminate their physical IT infrastructure and the associated costs. Cloud computing also enables organizations to deliver secure remote work capabilities to their people, more easily scale their data and apps, and take advantage of IoT.
Edge computing allows the capture, processing, and analysis of data at the farthest reaches of an organization's network: the "edge." This allows organizations and industries to work with urgent data in real time, sometimes without even needing to communicate with a primary datacenter, and often by only sending the most relevant data to the primary datacenter for faster processing. This spares primary computing resources like cloud networks from being glutted with irrelevant data, which lowers the latency for the entire network. It also reduces networking costs.
Consider an oil drilling rig operating in the middle of the ocean. Sensors that track information like drill depth, surface pressure, and fluid flow rate can help keep the machinery on a rig running smoothly and help keep workers and the environment safe. To do this without slowing down the network unnecessarily, the sensors send only the data about critical maintenance needs, equipment malfunctions, and worker safety details over the network, and this makes it possible to identify and react to issues in close to real time.
Fog computing allows data to be temporarily stored and analyzed in a compute layer between the cloud and the edge for cases where it's not possible to process edge data due to edge equipment compute limitations.
From the fog, relevant data can be sent to cloud servers for longer-term storage and future analysis and use. By not sending all of the edge device data to a central datacenter for processing, fog computing allows companies to reduce some of the load on their cloud servers, which helps to optimize IT efficiency.
For example, consider a building management company that uses smart devices to automate temperature control, ventilation, lighting, sprinklers, and fire and security alarms in all of its buildings. Rather than having these sensors constantly transmitting data to their main datacenter, the company has a server in each building's control room that manages immediate issues, and only sends aggregated data to the main datacenter when network traffic and compute resources have excess capacity. This fog computing layer allows the company to maximize its IT efficiency without sacrificing performance.
It's important to note that edge computing is not reliant on fog computing. Fog computing is simply an additional option to help companies gain more speed, performance, and efficiency in certain edge computing scenarios.
Edge computing use cases and examples
IoT devices and edge computing are rapidly transforming the way industries around the globe work with data. Following are some of the most notable uses for edge computing in business:
Branch offices. Smart devices and sensors reduce the number of resources needed to run a company's secondary offices. Consider internet-connected HVAC controls, sensors that detect when copiers require repairs, and security cameras. By sending only the most essential device alerts to a company's primary datacenter, edge computing helps prevent server congestion and lag time while greatly increasing response time to facility issues.
Manufacturing. Sensors on factory floors can be used to monitor equipment for routine maintenance issues and malfunctions, as well as keeping workers safe. In addition, smart equipment in factories and warehouses can increase productivity, reduce production costs, and provide quality control. And keeping data and analysis on the factory floor rather than sending it to a centralized datacenter can help avoid expensive and potentially dangerous delays.
Energy. Power and utility companies use IoT sensors and edge computing to increase efficiency, automate the power grid, simplify maintenance, and make up for shortfalls in network connectivity at remote locations. Utility towers, wind farms, oil rigs, and other remote energy sources can be equipped with IoT devices that are able to withstand harsh weather and other environmental challenges. These devices can process data at or near the energy site and send only the most relevant data to the main datacenter. In the oil and gas sectors, IoT sensors and edge computing provide essential real-time safety alerts that notify key personnel about necessary repairs and dangerous equipment malfunctions that could lead to explosions or other catastrophes.
Farming. Edge computing can help boost agricultural efficiency and yields. Weather-resistant IoT sensors and drones can help farmers monitor equipment temperature and performance; analyze soil, light, and other environmental data; optimize the amount of water and nutrients used on crops; and time harvests more efficiently. Edge computing makes using IoT technology more cost-effective even in remote locations where network connectivity is limited.
Retail. Large retailers often gather massive amounts of data throughout their individual stores. By using edge computing, retailers can extract richer business insights and react to them in real time. For example, retailers can collect data on customer foot traffic, track point-of-sale numbers, and monitor the success of promotional campaigns across all of their stores and use this local data to manage inventory more effectively and make faster, more informed business decisions.
Healthcare. The uses for edge computing in the healthcare sector are vast. Temperature sensors shipped with vaccines can help ensure that they maintain their integrity throughout the supply chain. At-home medical equipment like smart CPAP machines and heart monitors can collect patient data and send relevant information to a patient's doctor and healthcare network. Hospitals can better serve patients by using IoT technology to track patients' vital signs and to more accurately track the location of equipment like wheelchairs and gurneys.
Autonomous vehicles. There is almost no margin for error with self-driving cars, taxis, vans, and trucks. Edge computing makes it possible for them to respond instantly and correctly to traffic signals, road conditions, obstacles, pedestrians, and other vehicles in real time.
Edge computing services
As edge computing has grown toward widespread adoption, the types of related services to support its use has also grown. Today's edge computing services go far beyond just devices and networking to include solutions to:
- Run AI, analytics, and other business capabilities on IoT devices.
- Consolidate edge data at scale and eliminate data silos.
- Deploy, manage, and help secure edge workloads remotely.
- Optimize the costs of running edge solutions.
- Enable devices to react faster to local changes.
- Ensure that devices operate reliably after extended offline periods.
The latest solutions include services to help incorporate edge computing with common technologies like databases, operating systems, cybersecurity, blockchain ledgers, and infrastructure management, to name just a few.
Examples of Microsoft edge computing services:
Azure IoT Edge
Extend cloud intelligence and analytics to edge devices
Azure Stack Edge
Bring Azure compute, storage, and intelligence to the edge with Azure-managed devices
Azure FXT Edge Filer
Support HPC workloads with a hybrid storage optimization solution
Azure SQL Edge
Get real-time data insights for IoT servers, gateways, and devices
Accelerate edge intelligence from silicon to service
Azure Data Box
Quickly and cost-effectively move stored or in-flight data to Azure and edge compute
Azure Network Function Manager
Deploy and manage 5G and SD-WAN network functions on edge devices
Windows for IoT
Build intelligent edge solutions with enterprise-grade developer tools, support, and security
Avere vFXT for Azure
Run high-performance, file-based workloads in the cloud
Azure Front Door
Get fast, reliable, and more secure cloud content delivery with intelligent threat protection
Azure confidential ledger
Store unstructured metadata in a blockchain using a REST API managed service
Securely connect MCU-powered devices from the silicon to the cloud
A note on AI and analytics edge computing services
AI and analytics services for the edge are especially valuable for improving automation, productivity, maintenance, and safety. Here's just one example: Deploying predictive models to factory cameras can help detect quality control and safety issues. In this case, the solution triggers an alert and processes the data locally to execute an immediate action or sends it to the cloud for instant analysis before taking action.
Edge computing is a networking technology that enables devices in remote locations to process data and perform actions in real time. It works by minimizing network latency through processing most data at the "edge" of the network—such as by the device itself or by a nearby server—and only sending the most relevant data to the main datacenter for near-instant processing.
"Edge cloud computing" is another way of saying "edge computing"—the two terms mean the same thing: enabling devices in remote locations to process data and perform actions in real time by minimizing network latency.
Edge computing technology includes networking solutions and hardware to allow smart devices in remote or challenging environments to function without needing a full connection to a central network. Networking solutions include technologies like 5G and solutions that help to reduce latency by minimizing the amount of data sent across the network. Common edge devices include cameras, sensors, servers, processors, switches, and routers, which connect over the network to a central datacenter. In many cases, edge devices run AI locally and send only certain critical data to the primary datacenter for additional processing.
Edge computing is often used for places like factory floors, retail showrooms, shipping containers, hospitals, construction sites, energy grids, and farms—and even the International Space Station—where devices or sensors need to work in real time but have only limited connectivity to a primary datacenter. It allows businesses to do things like use sensors to make sure machinery is operating safely and efficiently, to detect when inventory is low on store shelves, to increase or reduce irrigation on farms based on soil moisture, and to spot when workers might be in danger.