According to a recent IDC report, it is projected that by 2026, 75 percent of organisations in the APAC region will depend on artificial intelligence (AI) to optimise asset efficiency and enhance product quality in diverse and distributed environments. In today’s digital economy, where customer expectations for seamless and personalised experiences are soaring, AI-powered networks play a pivotal role in delivering these experiences. With over 55 percent of IT leaders already facing resource constraints in network management, the question arises: how can AI-driven network automation elevate organisations’ network strategies?
CIO World Asia had the opportunity to interview Nick Harders, SASE Director, APJ, at HPE Aruba Networking. Nick will shed light on how organisations can harness AI for IT operations (AIOps) to unlock their network’s potential and enable next-generation digital experiences.
The Crucial Role of a CIO in Navigating AIOps Complexity
In the current landscape, modern Chief Information Officers (CIOs) are facing a unique and pivotal role as organisations embrace the integration of artificial intelligence (AI). The advent of generative AI has injected new enthusiasm into the realm of applied AI, sparking a transformation in traditional business practices.
According to the latest research conducted by HPE Aruba Networking, IT leaders in Singapore are wholeheartedly embracing emerging technology, with a striking 93 percent already utilising AI and machine learning (ML) models or planning to do so. As businesses increasingly rely on AI, CIOs are confronted with the challenge of effectively managing the associated risks and devising a comprehensive strategy for Artificial Intelligence for IT Operations (AIOps).
Given the substantial hurdles of integrating AI into existing systems, optimising these solutions at scale may necessitate significant changes in business processes and a reimagined approach to collaboration between AI teams and software engineering
The Crucial Emphasis on Preparation Over Technology in Achieving AIOps Success
Numerous organisations continue to rely on outdated networking hardware, including ageing wireless access points, switches, and obsolete WAN infrastructure, ill-suited to support the demands of modern AI-driven workflows. Despite the enduring value of data from legacy infrastructure, contemporary network management tools offer intricate telemetry, enabling more comprehensive monitoring, troubleshooting, and optimization insights and recommendations.
To unlock the potential of AIOps, it is imperative for business leaders to craft a well-defined implementation strategy and identify issues that AIOps can effectively address. For instance, in healthcare environments densely populated with Internet of Things (IoT) devices, network performance can deteriorate. A case in point is a hospital where patients and visitors reported subpar connectivity experiences despite the appearance of a robust Internet connection.
With the aid of AIOps, the IT team pinpointed that devices from adjacent buildings and the surrounding area were intermittently connecting to the hospital’s Wi-Fi as people passed by, severely hampering network performance. By isolating the root cause of the issue, the IT team effectively managed IoT traffic within the hospital premises, resulting in an 80 percent reduction in help desk calls and a significantly improved user experience.
This highlights the critical need for pinpointing specific challenges requiring resolution and ensuring that the infrastructure is primed for AIOps integration. Equipped with AIOps insights, IT teams can proactively deploy automated self-healing mechanisms, even from remote locations, to swiftly address and mitigate such issues.
The Impact of Outdated Data Strategies on AIOps Efficiency
AI and ML heavily rely on training data, a fact well-acknowledged by companies. Outdated approaches to data training can have a detrimental impact on the performance of AIOps. Data essentially serves as the lifeblood for AI/ML algorithms, enabling them to understand IT processes. When different departments or functions maintain isolated data repositories, it hampers the ability to gain a comprehensive view of business operations and complicates the process of extracting insights that span across the entire organization.
A research conducted by Aruba, aimed at exploring how enterprises perceive the potential of their networks, revealed that 96 percent of IT leaders in Singapore consider access to data as a critical factor in unlocking new revenue streams and services in the coming year. Furthermore, 93 percent of organisations in Singapore are actively seeking data-driven insights from their IT systems.
Surprisingly, even well-established businesses often fall short in providing AIOps-based analytics platforms with the requisite training data to enhance their performance and deliver relevant insights. Training data that is cluttered or contaminated with outliers introduces noise into the learning process of AI and ML algorithms. Consequently, these algorithms may struggle to identify patterns and make accurate predictions, resulting in suboptimal AIOps performance.
The Role of AI in Resolving Crucial Network Management Issues, Including Security and Automation
In the current 24/7 operational environment, maintaining network reliability has traditionally involved a deluge of help desk calls, extended troubleshooting sessions, on-site visits, and configuration attempts, often without a comprehensive understanding of the underlying IT issues. Surprisingly, only 32 percent of organisations report that their networks enable employees to work from anywhere, indicating the pressing need for networks to perform optimally, especially in the absence of physical IT personnel.
Numerous organisations grapple with fragmented systems that hinder the seamless access to data across various domains. AIOps addresses this challenge by consolidating application and business data into a unified platform and continuously monitoring network traffic data to deliver real-time insights into network performance, even in remote settings. Employing machine learning (ML) algorithms, AI leverages past network issues to predict maintenance requirements, conduct behavioural analysis of network users and devices, and significantly reduce troubleshooting time for network optimization and anomaly detection, including the identification of malicious activities.
The same principle extends to network security, where the quality and relevance of training data play a pivotal role in ensuring the effectiveness of AI-driven actions. AI equips IT teams to tackle network challenges, even in remote scenarios, safeguarding both performance and security.
Crafting an AI-First Approach for the Intelligent Edge
To effectively embrace an AI-first strategy for the intelligent edge, it’s crucial to take a comprehensive perspective, factoring in technical, operational, and strategic elements. The intelligent edge is designed to thrive in the mobile, IoT, and cloud-centric landscape, offering the potential to unlock substantial value in terms of responsiveness, user experience, and operational efficiency when executed thoughtfully.
According to Gartner, projections indicate that by 2025, over 50 percent of mission-critical data will originate and be processed beyond the confines of traditional data centres or cloud environments. Real-time AI inference capabilities become paramount for applications that demand low-latency responses. This is especially pertinent in sectors like autonomous vehicles and industrial automation, where immediate decision-making based on incoming data can be a game-changing advantage.
Successful implementation of an AI-first approach at the edge also hinges on algorithmic adaptability. While conventional AI models are robust, they can be resource-intensive. At the edge, there’s a demand for streamlined yet equally potent AI models. AI-powered systems provide insights into IoT traffic patterns and simplify device configurations, facilitating large-scale operational efficiency and automation. Additionally, interoperability plays a pivotal role in ensuring seamless interactions across a diverse array of devices and platforms.
Strategies of IT Decision Makers in Singapore for Approaching the Enterprise Network
Research findings illuminate a crucial moment within the digital transformation landscape. In Singapore, a remarkable 91 percent of organisations acknowledge the necessity of embracing an advanced level of digital transformation for their success. However, an equally notable 79 percent express concerns about their ability to keep up with this transformation.
This sentiment is not unique to Singapore; across all 21 surveyed markets, a significant 58 percent of IT leaders are calling for increased support from AI-powered network automation. They aim to reduce the time their IT teams spend reactively identifying and resolving issues.
In response to this urgent demand, the adoption of Network-as-a-Service (NaaS) is rapidly gaining momentum. 93 percent of IT leaders indicate that their organisations either have ongoing NaaS implementations or have imminent plans to deploy this innovative network approach within the next two years. NaaS offerings encompass integrated hardware, software, licences, and support services delivered in a flexible consumption or subscription-based model.
Networks managed under a NaaS framework undergo continuous monitoring for performance and security, potentially advancing toward a predictive environment where network telemetry drives machine learning-backed insights. As organisations venture into the realm of NaaS and AI-driven network automation, IT leaders can utilise these tools to unlock new opportunities for innovation and gain a competitive advantage in this digital landscape.
In conclusion, the integration of artificial intelligence and AIOps into network management is no longer a luxury but a necessity in the rapidly evolving digital landscape. The APAC region, with its growing reliance on AI-powered networks, is at the forefront of this transformative journey. As IT leaders, CIOs, and organizations embark on this path, they must prioritize careful preparation, data quality, and a comprehensive approach to unleash the true potential of AIOps. With the right strategies and innovative solutions like Network-as-a-Service, they can not only address current challenges but also embrace the future of network automation, making their mark in the era of intelligent, data-driven, and highly responsive digital experiences.