What is edge computing and how does it work?Edge computing refers to capturing, storing, processing and analysing data where it is generated, instead of in a central location. This significantly reduces what we call ‘latency’ – the delay caused by sending data to and from a distant location to be processed. In turn, this can deliver strong business benefits, including faster insights, improved response times and better bandwidth availability.

Edge computing helps you unlock the potential of the vast untapped data created by devices connected to the Internet of Things at different locations. You can uncover new business opportunities, increase operational efficiency and provide faster, more reliable experiences for your customers.

To gain the most value from all those devices, significant volumes of computation must move to the network edge. This allows connected devices to have more location awareness and handle activities autonomously. In banking, we’re already seeing this with new ‘smart’ ATMs – machines that banking customers can use to deposit funds, open accounts, transfer money, and other things that can usually only be done at a bank branch – and kiosk devices.

How does 5G relate to edge computing?

5G and edge computing both address challenges with bandwidth and latency, but in different ways. 5G standards have a much higher connection density, allowing networks to handle immense numbers of connected devices. These standards will increase bandwidth from 200mbps (4G) to 1gbps (5G) or higher, while also reducing the latency of moving data to the telecommunications network from 50–100 milliseconds to 1–4 milliseconds.
"Edge also works in tandem with distributed clouds – deployments of larger clouds in key locations for the business"

But data still needs to be sent to the cloud or data centre for processing, which can take several seconds. This is where 5G and edge become allies. Edge computing allows us to reduce roundtrips to the main data centres by processing the data where it is produced. And if we need to send processed data to a data centre, we can use the 5G network to do it quickly, even for large data streams.

Edge computing with 5G creates tremendous opportunities in many sectors. It brings computation and data storage closer to where data is generated, enabling better data control, reduced costs, faster insights and actions, and continuous operations. In fact, by 2025, 75% of enterprise data will be processed at the edge, compared to only 10% today, according to research firm Gartner.

How can artificial intelligence (AI) and edge computing work together?At IBM, we view AI and edge computing as symbiotic. Edge computing is all about processing and analysing data locally, enabling us to take immediate action. This means we need to bring more intelligence into the location by using AI and machine learning.

For instance, having the ability to recognise patterns or inspect video streams with AI is going to be a common edge computing approach, whether it’s to identify fraud during money transfers or to identify customers with special requirements entering a bank branch.

In what sectors has edge computing had big successes so far?Edge computing can be applied to nearly any industry or sector where there is a need for specific use cases, such as autonomous behaviour, visual inspection, or immediate reaction to events.

Some of the sectors and industries that have been most active in applying edge computing so far include financial services, manufacturing, telecommunication, retail, automotive, nautical and mining.

In financial services, for example, the applications for edge computing are wide-ranging – from advancing cyber and physical security and preventing fraud, to hyper-personalising customer services, whether it’s on a mobile investing platform or inside a bank branch.

We’re also seeing safety-focused applications of edge computing that will support the safe return to work as Covid-19 lockdowns end. For instance, at IBM we have a solution that helps clients manage facilities and optimise space allocation by using real-time data including WiFi, cameras, Bluetooth beacons and mobile phones. This data, collected in a way that is designed to preserve employees' privacy, enables managers to quickly reallocate spaces, designate no-go zones, arrange for cleaning and monitor crowding, social distancing and mask-wearing.

What are the benefits and the risks of edge computing to the financial services sector?Edge computing has the potential to benefit financial services institutions in several ways.

It can allow banks to create better customer experiences. For example, a device located at a bank branch could react to the context of individual customers. If someone entering the branch is in a condition where they could benefit from special assistance, such as a pregnant woman, staff could be alerted so they can offer personalised service. Or it could detect when a regular customer enters the branch, based on the Bluetooth signal from their mobile banking app, triggering a regular pattern of service.

Edge devices can also be used to monitor the length of teller queues, and trigger the opening of a new teller when the queue gets too long.

They can enhance customer relations by supporting applications like real-time language translation and more advanced chatbot services with audio and video functions.

Edge computing can also improve safety and security for banks. For example, it could allow cameras to detect when someone enters a bank branch with weapons or enable rapid identity verification in places with intermittent connectivity. In addition, edge allows for pattern recognition to help spot criminal activities, by using real-time video analytics from data that has been aggregated from multiple branches connected to the cloud.
"Transforming a centralised system into a distributed system while maintaining a holistic overview of the business isn’t a trivial task"

Edge solutions are also safer because they allow for workloads and data to be distributed so that you don’t have all the data in one location. For example, some data could be stored at a bank branch or on your personal mobile device.

When it comes to IT operations, edge computing allows systems for applications like ATMs and kiosks to be updated autonomously, wherever they are.  

Edge also works in tandem with distributed clouds – deployments of larger clouds in key locations for the business, such as a bank branch. This helps address data residency issues that financial services firms face when operating in multiple countries with different regulations. Edge computing can help keep workloads updated according to set policies to meet local regulations.

There are of course areas of risk in edge computing that financial services firms will have to manage. For example, distributed environments extend an organisation’s ‘attack surface’, meaning the locations where a cyber attack could be targeted. Proper measures need to be taken to embed security and safety into the activation of the edge locations and devices.

Management and operations processes also need to change and become fully automated, to avoid the need to have IT experts in every location with an edge device.

Transforming a centralised system into a distributed system while maintaining a holistic overview of the business isn’t a trivial task. It’s crucial to have a deep understanding of what kind of data the business needs and captures in different locations.

How can edge computing impact the experience of financial services customers, including vulnerable customers?We’re increasingly seeing banks deploying self-service locations and smart branches, where kiosks provide a wider range of customer services. There is huge potential for edge computing to support this trend and transform customer experiences.

For instance, using smart cameras with customer profiles allows customers to be more accurately identified and directed to the right branch employee. This can particularly benefit customers from vulnerable groups, such as people with different types of disabilities, while creating opportunities for bank staff to offer more personalised products and services.
"Edge-enabled cameras at the ATM could identify if someone is trying to view your pin"

To achieve this, branches will need to be equipped with modern devices and applications, which work even with poor connectivity. It requires an autonomous management system to ensure the correct applications are activated at the right times and that kiosks and ATM capabilities are updated automatically at scale.

The outcome for banks that get this right could be lower attrition and more loyal clients, who have been served faster and made to feel special. These elements could even form part of the key performance indicators banks use for net promoter scoring ­– measuring customer experience and predicting growth – to gauge the success of these new capabilities.

Edge computing can also improve the experience of customers using ATMs, including by making them safer. For instance, edge-enabled cameras at the ATM could identify if someone is trying to view your pin, or if someone is being forced to make a withdrawal against their will. The system could also shut off an ATM that has been tampered with from the banking network, helping to prevent fraud.

How do edge computing and data security work together? And what impact could this have for consumers and workers within financial services?Making sure applications are tamper-proof and deployed securely to protect data should be part of any edge computing strategy. At IBM we achieve this with an application manager solution that allows software to be deployed to hundreds of thousands of devices while keeping data secure. We ensure data security by handling data only where it is needed and distributing only data that is essential for an activity – all backed by IBM's leading encryption technology. 

In the context of a bank branch, this could mean that data relating to customer transactions only needs to be stored at their branch, or even on their personal device, rather than being sent to other locations for processing or storing. This cuts the risk of so-called ‘man in the middle’ cyber attacks, where data is intercepted by an attacker interjecting themselves into digital communications. But should any criminal behaviour be detected, for example at a branch ATM, we can send the data feed to the security centre for further investigation.
"Ultra-fast 5G connections paired with edge computing devices have the potential to extend financial inclusion in different populations"

More broadly, regulatory bodies, such as the FCA in the UK, require financial services firms to mitigate risk across their IT estates and digital supply chains. IBM launched the sector's first custom-made cloud platform to help solve this challenge. It offers the strongest commercially available encryption technology paired with confidential computing, which means our clients' data is only ever accessible to them. At the same time, it also enables firms to ensure automated compliance continuously, with regulatory compliance controls built into its core. 

Our cloud platform helps de-risk the entire financial services supply chain, because the banks, fintechs, and technology providers using the platform all adhere to a consistent set of standards. This also means banks can onboard technology vendors in a fraction of the usual time and consume more innovation, to the ultimate benefit of consumers. 

How will edge computing enable financial institutions to scale their level of operations, for example, by avoiding network delays?Edge computing will provide financial institutions with more reliable systems for their business because processes will be executed locally, and therefore much faster. This in turn makes it easier for businesses to operate across different geographies because they can perform IT operations according to a schedule that best suits a particular location.
About the expert

Dennis Lauwers is the distinguished engineer for hybrid cloud in Europe, Middle East and Africa, focusing on client leadership and bringing new hybrid cloud technologies to the market. Dennis joined IBM in 1997 and has held different positions within IBM, including system engineer, application and integration architect, and European technical leader for private cloud.

Edge computing also gives them the ability to introduce new functions quickly for their clients or install pop-up locations at specific events – driving new ways of doing business.

Edge computing allows for faster accessibility for mobile banking apps. Will this help with financial inclusion?Ultra-fast 5G connections paired with edge computing devices have the potential to extend financial inclusion in different populations. For example, it could facilitate next-generation chatbots with real-time translation and video functionality that would make online banking more accessible to people with limited language ability. And the faster, more reliable connections 5G offers can enable banks to operate in remote areas that have seen local branches close. Whether it’s through mobile or pop-up banks or more advanced mobile app-based services.  

Edge computing can also offer similar inclusion benefits with 4G networks, when 5G is not available or affordable. For example, delivering new user experiences and making banking services easier to use. 

What are the cost implications for using the technology? While there are costs involved in implementing edge computing, such as installing new devices and applications, it will also drive cost efficiencies. When you have to bring all of your data to a data centre for analysis, you have to upgrade the network and security technologies to handle this increasing workload annually, at significant cost. Edge computing avoids this expense by decentralising a company’s data, which also reduces the risk to business continuity from an outage. 

The benefits edge computing brings to a business, such as improving customer experiences and loyalty as well as enhancing security, should ultimately offset costs in the long term.