The Internet of Things heralds a day where almost any...

The Internet of Things heralds a day where almost any environment—whether it be a shop floor, a retail store or an entire city—can be made intelligent by sensing, analyzing and reacting to new information almost instantaneously.

It would seem natural to assume that the cloud, with its massive computing power, would be the “brains” of such an intelligent environment. But the task of collecting and processing data from countless devices in one central location quickly becomes too expensive or time-consuming, notes Frank Piller, a business technology expert at RWTH Aachen University. “That’s where edge processing comes in,” Prof. Piller says.

In edge computing, data is processed close to its source, whether in a device or other localized machine. Edge computing enables the IoT in real-world conditions where connectivity can falter and data transfer carries cost and security risks. For example, in a manufacturing environment, “Edge computing reduces the bandwidth and latency hurdles to advanced tasks such as image or video analytics, predictive maintenance and machine-learning-based applications,” says Srikanth Jayaraman, Technology Director, IoT WoRKS at HCLTech.

“It’s bringing the metaverse into the real world.”

Frank Pillerbusiness technology expert, RWTH Aachen University

Today’s sensors pack more computing power into smaller and cheaper packages, which enables more processing to be done on the edge. But building an intelligent environment requires more than stringing together a collection of smart devices. Intelligent Secure Edge from global technology major HCLTech, incorporates technologies such as Vision AI, Wi-Fi 6, 5G and collaboration platforms into scalable IoT solutions that maximize impact while minimizing cost and disruption.

Factories on the edge

The shop floor presents challenges to building a reliable network to access the cloud, notes Jayaraman. Heavy equipment on the shop floor can interfere with communication. Factories often use unique networking protocols and legacy operating systems, and must be physically isolated from the Internet to prevent security breaches. Processing data on the edge can help manufacturers overcome such obstacles.

To take one example: A major truck manufacturer ensures the quality of its paint jobs by monitoring the temperature, pressure and humidity in the box where the painting occurs. When an inspector notices a potential quality issue, he or she takes an image for automated inspection. But passing huge images to and from the cloud uses massive bandwidth and incurs cloud utilization charges. Even a brief loss of connectivity could halt the production line.

Working with HCLTech, the manufacturer developed an edge gateway that processes the data first. The monitoring is in real-time; only anomalous data is sent directly to the cloud for analysis, while the rest is archived. The gateway runs even when the cloud connection is broken. The result is better quality, a more reliable process and a better analysis of how environmental factors affect the painting.

The cloud carries advantages such as scalability, enhanced collaboration and IT savings. As companies move more applications to the cloud, edge computing ensures that applications that benefit from localization continue to run smoothly, Jayaraman says. Edge computing ensures that the right data is processed in the right place and right time. It eliminates the security risk of sending data to the cloud and helps with compliance to data-sovereignty laws.

The edge also enables new capabilities. For example, HCLTech helped a train operator develop a system in which a camera device mounted on each train captures pictures of the tracks during movement and supplies them to an edge device at each station. If a crack is detected, the images are pushed to the cloud, which analyzes the data, takes immediate action if needed (such as telling the next train to slow down) and alerts maintenance staff.

Edge for smarter cities

Edge computing also plays a crucial role in smart cities, which require deployment of countless sensors for collection and analysis of data across wide areas. In the borough of a major Latin American city, a smart edge-based system called ISE for Smart Cities uses AI-enabled cameras and microphones to detect persons of interest, monitor traffic, and stay alert for signs of weapons and smoke.

ISE for Smart Cities has been helpful in combatting crime. Each resident can also request a security kit installed at their front door. The feed from the kit is analyzed at the local police station. A social app enables citizens to report security issues or ask for a police patrol. One executive credited ISE for Smart Cities with helping the municipality reduce high-impact crimes by 46%.

Instead of digging up the streets to install a network of cables to connect the cameras to the cloud, HCLTech and its partners process much of the data at the source and send it wirelessly through a Wi-Fi 6 mesh network. “The camera device, in tandem with the edge layer, is analyzing whether action needs to be taken in real time, and what information must be sent where,” says Sridhar Negamanthan, global vice president and global head business innovation, HCLTech. This reduces latency problems and bandwidth requirements.

The system can be deployed rapidly thanks to a modular design; a kit of 50 cameras covers four to six city blocks. “Within a week, the kit can be installed by a two-person crew without trenching,” says Ashay Chaudhary, Principal Technology Architect, IoT WoRKS at HCLTech. Best of all, networks like these can be used for a variety of other smart city applications, from environmental quality reporting and asset monitoring to disaster response.

A “physical metaverse” for retail

Busy stores help contribute to a thriving urban environment, and edge computing is helping brick-and-mortar retailers keep up with their online competitors. The online trade has held an advantage over physical stores in terms of customization and personalization, but that is changing. For example, a customer walking into one major chain’s beverage stores in Latin America sees digital advertising that is instantly customized to his or her age and demographic. AI-enabled cameras analyze the customer, supply the ads and measures customer engagement. The solution from HCLTech and its partners has resulted in a 25% increase in impulse buying and a 35% increase in sales revenue.

Customer-specific information is never stored and never needs to go to the cloud. This helps ensure that the right ad is displayed quickly. It also helps address privacy concerns.
“It’s bringing the metaverse into the real world,” says Prof. Piller.

The uses for intelligent edge are potentially boundless, but only when it is part of a well-designed IoT solution. These solutions often involve bringing together many different technologies to serve organizations facing a multitude of different demands. “I don’t believe there is one vendor who has all solutions under one roof,” says Mr. Chaudhary. “That’s where the importance of a system integrator comes in, because we design it to be future-ready.”