How edge computing and 5G can drive business applications

Over the past decade, advances in cloud computing have led to a centralized approach to system operations and management, while the development of mobile computing, the Internet of Things (IoT), and SaaS have driven computing toward a distributed architecture. With the introduction of edge computing and 5G technologies, companies are now trying to take advantage of both approaches while improving application performance.

The hype around edge and 5G tends to focus on innovation. Experts say cutting-edge applications such as driverless cars, virtual or augmented reality (VR/AR) and robotics are not limited to these applications and offer a wealth of opportunities for IT professionals.

How edge computing handles latency
Over the past few years, enterprises have benefited from cloud computing by centralizing resources in data centers owned by cloud providers. In-house data centers focus on avoiding capital expenditures and saving money on management costs. However, centralization has led to performance issues in handling Internet edge endpoints, including IoT sensors/devices and mobile devices.
Even now, smartphones are smart computers that fit nicely in your pocket, but they still lack the ability to do a lot of processing in the cloud. "Why can't you put all the intelligence at the end? In other words, why can't your smartphone do that?" asked Mahadev Satyanarayan, a computer science professor at Carnegie Mellon University.
In response to that question, he said, “The answer is to do the computation you want to do; you need far more computational resources than you can carry on a smartphone.” He added, “If you think about the video camera on a smartphone, it’s very lightweight. But if you want to do real-time video analytics on it, you can’t do that with the computer on the phone today — you’re sending the data to the cloud, and that’s the problem.”
An influential 2009 IEEE pervasive computing paper (co-authored by Satyanarayanan) outlined a solution to this problem: using virtual machine-based “cloudlets” for mobile computing. In other words, clustering small pieces of data at the edge of the network, close to where their processing power is needed.
Satyanarayanan explained that on 4G LTE networks, the round-trip time between a smartphone and a cell tower is about 12 to 15 milliseconds, and can be longer depending on legacy systems and other factors. However, when you try to connect a smartphone to a data center, it can take 100 to 500 milliseconds. In some cases, it can even take a full second.
The appeal of edge computing is that it reduces the tail of the distribution.
Data transfer speeds on 5G networks
The concept of "mobile intelligence" didn't become popular until four years ago. That's when telecom companies realized the need for 5G speeds and began making plans for 5G wireless.
While data transmission time on 4G networks is between 12 and 15 milliseconds, vendors are touting 2 to 3 millisecond latency levels for 5G, however, the round-trip time from a distant data center can still take around 100 to 500 milliseconds. "It doesn't make sense if you have to go back to a data center in another part of the country or the world, even if it's just a matter of milliseconds," Satyanarayanan said.
Dave McCarthy agrees with Satyanarayanan, research director at IDC Edge Strategies: “5G itself reduces network latency between mobile towers and endpoints, but it doesn’t promote proximity to data centers, which can cause problems for latency-sensitive applications.”
He added, “By deploying edge computing into 5G networks, it reduces this physical distance and greatly improves response times.” This makes edge computing critical to the rollout of new mobile edge computing (MEC) services and 5G networks.
Experts say it's critical to understand that 5G and edge computing are not tightly coupled. 5G networks require edge computing technology to succeed; edge computing runs on different networks, such as 4G LTE, Wi-Fi and other network types.
How do 5G and edge drive business applications?
When you combine 5G speeds with the processing power of edge computing, it's natural to focus on applications that require low latency. That's why early use cases tend to involve VR/AR, robotics, and AI, which require computing resources to make decisions within seconds. However, a variety of business applications have the potential to benefit from 5G and edge.
“Many applications that already exist in the local edge can essentially ‘move’ or take advantage of mobile edge computing,” said Dalib Adib, edge computing practice lead at STL Partners.
“There are so many use cases using video, AI and IoT,” he added.
Experts cite numerous use cases for edge computing in the enterprise, including:

• Real-time process optimization in production facilities. Data generated by smart connected equipment can not only dynamically adjust calibration settings, but also improve yields and reduce defects.

• Condition-based monitoring – using IoT devices/sensors to check specific parameters on a machine to ensure it is functioning properly.

• Businesses with capital-intensive assets in industries such as manufacturing and oil and gas, using edge and 5G for repair and maintenance. This includes AR/VR applications to guide technicians through repairs, using advanced analytics to help identify potential defects or products in need of maintenance, and drones to conduct visual inspections of bridges, buildings, or rail lines.

• Surveillance video analytics, for example, using real-time processing to determine whether an individual entering a building is an employee or visitor and to ensure the employee’s identity.

• Video analytics to provide real-time recommendations to law enforcement decision-makers during emergency situations.

• Telemedicine applications in healthcare – using video and analytics when diagnosing patients, or for remote patient monitoring.
Satyanarayanan foresees the development of edge-native applications that take advantage of edge computing, such as bandwidth scalability and low latency. These applications are likely to drive growth in edge computing and 5G networks.