By 2025, 50% of enterprise data will be created and processed at the edge, compared to the present 10%. Courtesy: L&T Technology Services
- Learn how many technologies including 5G, Multi-Access-Edge Compute (MEC), miniaturization and virtualization of network infrastructure and reservable spectrum for industrial use are coming together to enable breakthrough new applications.
- Learn about edge computing use cases in health care, automotive, manufacturing, asset management and retail.
Edge computing insights
- Convergence of edge computing and 5G enables unmatched accessibility to data and applications by bringing down latency and optimizing service delivery at the network edge.
- Connectivity combined with advanced computation delivers immense potential to many industrial stakeholders.
- There are several existing edge uses cases that are expected to improve with a higher rollout of 5G in industries including health care, automotive, manufacturing, asset management and retail.
It is impossible to exaggerate the role that edge computing is poised to play in creating and processing data. In fact, by 2025, 50% of enterprise data will be created and processed at the edge, compared to the present 10%.
At the same time, investments in 5G infrastructure are accelerating rapidly as deployments pick up pace. McKinsey predicts that by 2025, telcos will devote over $600 billion toward 5G infrastructure.
It is clear that edge computing and 5G are changing the way we utilize and harness data, be it across manufacturing operations, asset management, omnichannel operations, smart grids, freight monitoring, intelligent transportation systems, public safety or emergency response systems. However, since they focus on different areas of data activity, their advancement begs the question: “What happens if they are combined?”
The short answer is that this convergence enables exciting new applications for industries that so far weren’t possible.
Convergence of edge computing and 5G enables unmatched accessibility to data and applications by bringing down latency and optimizing service delivery at the network edge. This is vital to emerging new-age applications including autonomous vehicles, automated robotics, enhanced safety and augmented/virtual reality.
The potential of synergy: Unleashing the power of convergence
Edge computing, combined with 5G, can elevate digital experiences, enhance performance, promote data security and enable uninterrupted operations in multiple industries. It empowers an unprecedented level of data accessibility along with reliable and secure access to two-way communications. Connectivity combined with advanced computation delivers immense potential to many industrial stakeholders.
Edge computing brings data storage and computation nearer to where data is created by places, things and people. Edge data centers create localized processing areas that collect and analyze data locally, lowering the latency that typically occurs if centralized cloud application were to be used. This latency reduction has two components—network latency and application-level latency. Network latency is the round-trip time taken by packets to traverse the network and back minus the compute time that an application may take. The 5G standard itself enables lower network latency especially as Release 16 with Ultra Low Latency Connectivity (URLLC) networks get deployed. Application latency is the time taken by the compute itself. Through tight coupling of the network and compute using Multi-Access-Edge Compute (MEC) software platforms running on the edge servers that communicate seamlessly with the 5G network elements, application latency can be reduced. For latency critical applications, both of these elements working together are required to deliver low enough latency for computer vision processing and immersive applications.
It has taken technology advances in many different areas to arrive at these capabilities. As mentioned above, the 3GPP 5G standard and MEC are two of these advances. Other key advances include the miniaturization and virtualization of network infrastructure, increases in edge computing capability, reservable spectrum availability for industrial use and, in the case of artificial intelligence (AI) and machine learning applications like computer vision, the progress in speed and accuracy now possible in object detection and classification.
It used to be that cellular base-stations were large and expensive, significantly more so than Wi-Fi network infrastructure. Device side cellular modem chips were also much more expensive than Wi-Fi client chips and devices. With 5G, this gap has narrowed considerably, to the point where 5G small cells are the size of a pizza box and total cost of ownership (TCO) of the network is lower for private 5G TCO than Wi-Fi in many large area industrial environments. Moreover, much of the 5G network functionality is now implemented in software as opposed to needing dedicated chips and hardware thus making it possible to update the capabilities of the network and apply security patches remotely.
Edge compute servers have undergone a similar transformation. Whereas in the past compute racks were large and bulky, a server the size of a briefcase can now run powerful enterprise applications.
Spectrum is another major leap forward. In a few countries including the U.S., UK and Germany, spectrum is reservable for use easily and directly by industries for exclusive use in their locations without the need to go through the complexities of reserving mobile operator spectrum. As opposed to using unlicensed Wi-Fi spectrum for mission critical connectivity, using reserved spectrum allows industries to avoid production line stops, safety incidents and the like, thus reducing the risk to the business.
Leaps and bounds in AI/ML advancements over the last few years have also opened up immense business opportunities to improve operational efficiency, safety, quality and cost. 5G cameras can capture video streams to monitor product quality coming through production lines. 5G connected drones can inspect solar, wind, oil refinery, train or power lines for defects. Warehouse robots can zip past each other and past people working in close proximity safely by using real-time computer vision applications that run on AI and ML. However, a high throughput, low latency 5G network and edge computing is a requirement for these AI/ML applications to deliver results in real time. Moreover, all this video data can be made secure by being confined within the industrial location and geographical area of operation, thus reducing security vulnerability for the enterprise.
Cutting-edge immersive experiences supported by edge computing
Edge enables enterprises to enhance and improve the ways they use and manage physical assets, allowing them to develop unique interactive experiences for their customers. Today, brands are leveraging edge computing to process and respond to customer engagements at unparalleled speeds, ushering in unparalleled innovation in CX. Also, as edge computing does not depend on internet connectivity, enterprises can support uninterrupted CX, regardless of unstable connections or server outages. The requirements for open radio access networks (ORANs) will also see a huge boost if the speed and low latency of 5G comes to fruition via robust implementations.
With edge computing, brands can deliver hyper-personalized and omnichannel CX by giving access to services that work in parallel with edge devices. As edge brings down data latency related to cloud computing, it opens newer avenues to deliver better customer service. It provides customers with abundant opportunities to interact with brands in real time, across multiple channels, allowing them to control their own narratives.
Use cases for 5G with edge computing
There are several existing edge use cases that are expected to improve with a higher rollout of 5G.
Health care: Robot-assisted surgery procedures can make the process smoother for surgeons, and the surgery shorter and less invasive for patients. In this context, edge computing results in multiple minor changes that add up to create a major impact. Incision sizes are reduced and the surgeon does not need to stand, having an optimal view of the site. They can also leverage controls that are more intuitive and natural.
Automotive: Edge computing empowers autonomous vehicles with increased safety, faster decision-making, improved reliability, optimized bandwidth utilization and enhanced privacy. Edge accelerates safety in these vehicles through rapid sensor data analysis and identifying obstacles, if any. Its decentralized architecture elevates reliability, enabling vehicles to operate even during network disruptions. Local data processing elevates data privacy and security by preventing transmission to remote servers. Edge computing also optimizes bandwidth usage by limiting the volume of data sent to the cloud.
Manufacturing: Edge computing allows manufacturers to introduce automation in their supply chain and factory floor through machine-to-machine communication and advanced robotics, all of which are nearer to the data source. Instead of transferring data to a server for analysis and response, edge computing enables tasks like pipe flow monitoring, machine cycle tracking and fatigue detection in sheet metal to be performed locally. This approach reduces latency, enabling speedy analysis and corrective actions.
Asset management: Edge computing can enable remote asset management by supporting real-time monitoring and decision-making nearer to the assets. With edge computing, organizations can collect and process data locally, thereby improving response times, maintenance, uptime and asset utilization. Edge computing also minimizes enterprises’ reliance on constant network connectivity and empowers them to optimize operations and maximize productivity.
Retail: In retail, edge computing enables enterprises to propel real-time marketing to enhance CX. As edge computing can swiftly react to customer inputs, brands can create hyper-personalized experiences and drive revenues and loyalty. A major application of edge in retail is frictionless store checkout. An in-store edge network can process data collected by in-store cameras using AI, which has been trained to recognize store items, enabling customers to go out of the store past a booth that charges their accounts automatically. This, of course, eliminates the need to wait in line.
Benefits of edge computing
With 5G, edge computing generates opportunities for new platforms, experiences and products in every industry. By using the computational power of edge devices, networks and gateways, enterprises can take advantage of continuous delivery and the principles of robust resource allocation that are intrinsic to cloud computing.
Modern-day businesses can also virtualize the cloud beyond the data center. Workloads generated in the cloud, including modern versions of AI and analytics, can be migrated toward the edge.
Edge computing supports:
Improved data control and reduced costs by minimizing vulnerabilities and reducing data transport to central hubs
Swifter insights and actions by leveraging increased sources of data and processing the data at the edge
Uninterrupted operations by supporting systems that run on their own even without an internet connection to reduce costs and disruptions
Welcome to the future: multi-access edge computing
Edge computing as the next phase of cloud computing helps to localize data for users, thereby cutting down on network load. Multi-access edge computing or MEC represents a major step ahead that shows how cloud computing can be used for faster communication and networking. MEC provides unparalleled processing and network speed when used in conjunction with 5G, with high bandwidth data transfers and low latency connections.
The convergence of 5G and edge computing will boost business in all industries and transform the way businesses operate and how people work. It will empower brands to perform real-time marketing and deliver hyper-personalization, recommendations, walkout shopping and various other advanced practices to enhance customer experience while optimizing the costs of transferring data. On the shop floor and at an enterprise level, it will impact all areas of industrial grade connectivity from mechanization to the assembly line and automation, leading to hyper personalization as well as real time corrective actions.
Adopting an open, hybrid, multi-cloud architecture will ensure that enterprises can deliver innovative connected experiences using data, irrespective of whether it operates in a public or private cloud, or on a centralized or on-premises data center.
Keywords: Edge Computing, 5G
What edge computing use cases can you most benefit from?