Smart edge: the future of cloud computing?
The intelligent edge refers to the collection and analysis of data where the data is generated. It can be the point where a user such as a mobile worker does their job, or an IoT device connected to industrial equipment generates data. Data is collected, processed, and analyzed at this remote point, which means intelligence sits at the “edge” of the organization’s IT architecture. This “decentralization” of data storage, processing and intelligence capabilities moves much of the workload away from the central processing hub, data center or cloud.
Edge computing is a powerful enabler when injected with intelligence to enhance its disruptive potential and enrich the value of data captured at its proverbial source, says Scott Cowling, director of network solutions at BT. “The sheer volume of data created through the IoT, where information is streamed from sensors to optimize operations, makes this intelligence a business imperative.”
For Hanno Brink, machine learning engineer at Synthesis, the technical challenge that the intelligent edge tries to solve is the efficient use of resources, the increase in robustness and the reduction of the costs of implementing and managing this technology. .
“Many benefits could be unlocked if this technology is applied correctly. One is reducing latency between when data is collected and processed, which has applications such as predictive maintenance, fleet maintenance and many others where real-time visibility and responsiveness would help reduce operating costs or improve service delivery.
“Another solution would be for smart applications to be delivered to customers where limited bandwidth or connectivity previously prevented them, or where customer data has not been easily collected and operationalized. This technology could also be used to collect data that was not previously accessible, due to the cost of implementation, and to provide new customer experiences or improve current service delivery by using intelligence where the customer interacts with it.
Forrester’s 2021 report “Building Integrated Technology Platforms to Accelerate Growth and Agility” tells IT leaders, “At this stage of technology maturity, innovate with technology to serve customers across the lifecycle of the customer, by mastering the systems of commitment and insight. »
This is precisely what the intelligent edge allows us to do, says Varsha Ramesar, head of data management and analytics at iOCO. “From a technology perspective, the intelligent edge has several benefits, including reduced dependence on network performance, and enables the business to increase its bottom line by reducing overhead. What organizations need to keep in mind, however, is that the true value of the intelligent edge lies in how it can extend and amplify an organization’s ability to sense and react with greater speed and agility – whether in the context of predictive maintenance or service.”
Navinder Singh, managing director of In2IT Technologies, explains that edge computing has emerged in recent years thanks to a few tech companies that have developed the technology. When the intelligent cloud arrived, it was a major innovation because it helped overcome some of the fundamental problems of cloud computing such as latency, bandwidth requirements and cost containment. Over the past year, cloud providers have accelerated intelligent cloud offerings, where data that sits at the edge with built-in intelligence can still interact with and relay to and from the cloud.
Moreover, he says investing in the periphery has changed its potential and scope. “If we look at the intelligence embedded in edge devices, we see advances that align with cloud strategies and one of the most challenging elements: integration with the cloud. Intelligence creates seamless cloud application integration via AI. For example, companies are often required to integrate their CRM and ERP solutions with cloud service providers, but there is a lot of effort required from resources for these applications to achieve seamless integration. The AI collects information from the edge applications and as the onboarding requirements are pre-defined, it greatly simplifies the onboarding process and therefore reduces costs. It also eliminates bottlenecks where processing and analysis have been done centrally with data alignment across multiple applications. The treatment becomes faster and more efficient.
Bringing data and compute closer to the edge also means that this infrastructure moves away from your control and becomes much more exposed.
Hanno Brink, Synthesis
However, like any other technology that is, relatively speaking, in its infancy, the intelligent edge is not without its challenges. According to Brink, these include creating more efficient hardware, creating more computationally efficient AI, finding more efficient ways to manage limited bandwidth and storage resources, securing the edge, distribute machine learning to the edge to maintain privacy and overcome the myriad challenges presented by implementation. systems that interact with the real world.
Various proof-of-concept and proof-of-value projects have proven the ROI and business benefits of intelligent edge deployments that bring AI and ML to edge environments, adds Ramesar. However, the challenge now is how to scale these deployments to hundreds or thousands of sites so that organizations can take full advantage of the critical data they generate at the edge. Before companies can take advantage of the benefits of the edge and embark on their industrial digital transformation, they need to consider their data. Data volume and speed are increasing astronomically, and availability is key. To support these applications and use cases, sensors and associated contextual data must be ingested, processed, and analyzed in the right place, at the right time, and delivered to the right people.
For Cowling, one of the biggest concerns at the edge and one of the biggest barriers to deployment is cybersecurity. Increasingly, the edge is becoming a point of convergence between two worlds: operational technology (OT), including the industrial systems that run equipment in factories, refineries, and mines, and computing. Industry 4.0 solutions such as predictive maintenance need data from both worlds, such as SCADA from OT and ERP from IT. air gaps. “Once you join the dots, critical processes become vulnerable because they run on old proprietary software, with poor password protection, limited patches and no authentication. Identifying and mitigating vulnerabilities therefore become a major focus.
When it comes to securing the intelligent edge, Brink says the most crucial component of any edge application is ensuring your application is secure from the device, to the cloud, and back. “Bringing data and compute closer to the edge also means that this infrastructure moves away from your control and becomes more exposed. Any such app should be designed with security in mind.
“Beyond data security, physical security and reliability also become factors that need to be considered and properly balanced with functionality. In addition, secure physical devices must be used, and these devices must be robust against the extreme environments in which they may operate.
Organizations must ensure that systems can scale effectively while ensuring consistent security implementation, and must define an organizational root of trust. It is a way for edge devices to authenticate themselves to the enterprise and prevent impersonation of privileged systems and abuse of their access.
Security and privacy risks can be reduced by limiting data flows between the collection point and central infrastructure, especially when those flows occur over the public internet, Cowling adds. Using the intelligent edge helps businesses comply with national data protection laws. It keeps sensitive data in-device, anonymizing and analyzing at source rather than sending identifiable information to the cloud.
The Secure Access Service Edge, or SASE, is also essential for bringing together network connectivity and security into a single, policy-based service that provides consistent, centrally managed end-to-end access and security. the other. SASE also supports a zero-trust approach to the cloud and underlying infrastructure, which means sessions are protected regardless of where the edge device connects from.
As we move closer to cloud and hybrid cloud environments, the benefits offered by the intelligent edge make it a compelling technology purchase decision.
Navinder Singh, In2IT
Securing the edge is not only daunting, it can seem downright impossible, given the unprecedented number of devices on the network, generating data every second of the day, all of which must be ingested, transformed and analyzed by computing platforms in nature. , and all must be locked.
“Surprisingly, many IT professionals think of security as securing perimeters and implementing robust access control. However, security in today’s edge and cloud era is so much more complicated. With the explosion of IoT, and Industrial IoT in particular, the attack surface has increased, as have the number of attack vectors. Because edge computing is a distributed model, its security concerns are very different from a centralized model. To reduce costs and speed deployment, many edge devices do not natively encrypt data, and IT managers need a security framework in place before large-scale deployment of edge projects,” says Ramesar.
Ultimately, as we move closer to cloud and hybrid cloud environments, the benefits offered by the intelligent edge make it a compelling technology buying decision, Singh says. It helps streamline business processes, align data across multiple applications, facilitate integrations, and more. It’s also about business optimizing operations and gaining efficiencies, and in the era of “always on” and “instant gratification,” the smart advantage is the way to go.
* This feature was first published in ITWeb’s May edition genius idea magazine.