This Industry Viewpoint was authored by Kailem Anderson, Vice President of Portfolio and Engineering, Blue Planet, a division of Ciena
You can’t go far in the tech industry these days without coming across a conversation about the future of artificial intelligence (AI). This conversation has been top of mind in the telecom industry for a while, and for good reason – network operators desperately want and need AI to help simplify their complex network operations especially when it comes to Internet of Things (IoT).
IoT’s role is to provide more real-time contextual knowledge that can assist in more intelligent decision making. While there are many use cases for IoT, a handful are evolving faster than others such as automotive, manufacturing and retail thanks to the cutting edge technologies providing the support, speed and resiliency they need. According to IDC projections, IoT devices will generate 79.4 ZB of data in 2025.
One of the most important elements for these use cases is the ability to provide just-in-time insights for appropriate actions to be taken. This ability requires that the intelligence be distributed and either put closer to the edge or on-device. The convergence of IoT and AI is known as ‘edge AI’ – the use of AI within IoT at end points such as sensors, devices, and connection points. For example, manufacturers leverage IoT to understand the health of their machines in the production line and then use AI for predictive maintenance and to ensure less disruptions to the assembly line.
AI in the networking world today is a technology that can augment network analytics to help drive business and operational efficiency. It enables the next generation of highly intelligent networks—a network capable of dynamically self-configuring or self-optimizing based on changing network conditions. AI will play a critical role in making decisions based on massive amounts of data and many infrastructure and operations teams will use AI-augmented automation in large enterprises. With the help of edge AI, inferences and insights can be provided much quicker and ultimately have a positive impact on costs – and a cost-benefit analysis can certainly help ensure the investment is a worthy one to make.
The best way to help get started with edge AI is to leverage a distributed AI analytics platform to support business-critical projects.
Distributed Data Collection for Immediate and Long-Term Decision Making
Crucial to edge AI is a platform that supports a vast array of analytical tools that can help provide inferences and insights closer to the edge, while also using the same data at the core for other use cases that are not mission critical but provide longer term analysis. Data collection at the edge will avoid adding overhead and delay since many IoT mission critical use cases require instant analytics. Data aggregation and analysis at the core is important to support for key tactical and strategic business decisions. Being able to correlate information and multisource data in this way will help achieve the right intent or outcome. Otherwise the insights are left to human interpretation which will add significantly more time in the decision-making process.
A picture is worth a thousand words as we all know. Although AI is getting better at inferring intent and actions, there is still progress to be made in communicating these in a meaningful way to the operator. A robust tool that includes graphs, charts and other visuals that help derive meaningful insights will reduce time to action. It is also important for the tool to be interactive, enabling operators to drill-down to more information and correlate data as needed.
Integrating With Other Systems
In today’s highly connected world, we use best-of-breed tools which do fewer things better. However, each of these tools independently are not necessarily the best in understanding intent. Integration is important not only to understand the big picture but also to make sure other tools are up to date. To make integration with other OSS/BSS systems simple and easy, the platform needs to provide an open, programmable interface and standard Rest APIs.
Supporting Closed-Loop Automation
Platforms need the ability to take actions on the network, or “close loops,” by leveraging policy engines in combination with an intent-based orchestration solution for more complex workflow automation based on the insights provided by AI. Closed-loop automation continuously assesses real-time network conditions, traffic demands, and resource availability to determine the best placement of traffic for optimal service quality and resource utilization. The platform must also support human oversight and control over any automation that is being programmed into the network to avoid any unintended outcomes.
For organizations that invest in edge AI, those ripples could turn into a tidal wave. IoT is a critical factor to consider when network operators and enterprises are building networks. With AI at the edge, meeting the rising demands of end users today for 5G, IoT and other next gen services will be faster, more seamless and affordable.
If you haven't already, please take our Reader Survey! Just 3 questions to help us better understand who is reading Telecom Ramblings so we can serve you better!Categories: Artificial Intelligence · Industry Viewpoint