How do AI and autonomy play into the space of networking?

You can also listen to this podcast on iono.fm here.

JEANETTE CLARK: ‘AI’, ‘autonomous’, ‘self-operating’ and ‘sentient’ – once these were buzzwords in any self-respecting science fiction movie, but today these terms are commonplace in discussions that centre on the technology that will enable our future.

But how do AI [artificial intelligence] and autonomy play into the space of networking? This is a fundamental layer in digital transformation, and we have a general idea of what is meant when we say autonomous-driving cars.

But what do we mean when we talk about autonomous networks or networking? Today we are posing these questions to Shamith Maharaj, executive for Data Networks at BCX. Welcome, Shamith.

SHAMITH MAHARAJ: Good afternoon, Jeanette, and thanks for having me.

JEANETTE CLARK: Let’s start at the beginning. What exactly is an ‘autonomous network’ and how far beyond basic automation do these networks go?

SHAMITH MAHARAJ: Jeanette, an autonomous network is a network that runs with minimal to no human intervention. It’s able to configure, monitor, and maintain itself independently.

JEANETTE CLARK: What would be some of those characteristics of autonomous networks, specifically in the African business landscape and setting?

SHAMITH MAHARAJ: That’s a great question, Jeanette. As you know, autonomous networks [have] existed for a while across the world from an African business landscape perspective. Some of the key characteristics would include the zero-touch provisioning, especially in hard-to-reach locations. An autonomous network has the ability to scale, and it’s fully application-aware, delivering an improved user experience.

JEANETTE CLARK: Could you perhaps give me an example of how an autonomous network already works in an African business case, just so we can make it very understandable to all our listeners?

SHAMITH MAHARAJ: Yes, it’s a good question, Jeanette. Looking at the African landscape, one of the key benefits or areas would be taking consumer products to a market when they don’t exist in your store – so giving consumers the ability to shop when they want to, ensuring that their needs are satisfied and they can secure their products as quickly as possible.

Now the network makes the decision in terms of identifying where the product is available, and how quickly it can get from the store to the consumer.

JEANETTE CLARK: Can you give me an example of how this works already, maybe in an area where BCX is involved?

SHAMITH MAHARAJ: We can, in terms of taking our products and services to new geographies and into new markets. We have service catalogues that are [readily] available to our customers and to the market. Expanding this to a new geographic footprint allows a larger audience to make use of our services, [showing] that we have provision in South Africa, as well as leveraging the partner ecosystems we have on a global scale.

JEANETTE CLARK: Can you take me through some of the capabilities that an autonomous network enables?

SHAMITH MAHARAJ: There are quite a few. I’ll touch on some of the key ones. With an autonomous network – it being self-driven, learning the user patterns – some of the key ones that we’ve seen to date have been track-and-tracing.

Making use of IP [Internet Protocol] cameras on the street we can help fight crime. In terms of the Covid pandemic that we just came out of, this would’ve been a key enabler to accelerate the tracking-and-tracing process.

On a more important note, it gives patients, from a medical perspective, a quicker reach to doctors in various geographies. Once symptoms are reported we can ensure that the correct medical support and advice is provided timeously.

What we’ve also seen, Jeanette, in terms of the manufacturing fields, is the next level of robotics being powered up using autonomous networks.

Another key one in our day-to-day lives includes facial recognition, especially at schools and at workplaces where we have stringent access control. This has been one of the key capabilities that autonomous networks [is] enabling and driving into the markets.

JEANETTE CLARK: Apologies for my ignorance on the topic, but if I, for example, use facial recognition, what makes it different from what we have known in the past, just as an automated network that maybe uses facial recognition? What is the key benefit from autonomous networks that is added into the mix?

SHAMITH MAHARAJ: That’s a great question, Jeanette. I think the answer there lies with the big data that’s being interrogated very, very quickly. So, with the processes built into the autonomous network, this allows the network to then become more self-aware, understanding what controls need to be applied in a specific situation or in a specific scenario. So I’ll give you a real-life experience.

In terms of a school, ensuring that learners at a specific school have access only to certain facilities in terms of presence. And in terms of managing the access controls of various areas of a prison, ensuring that lockdown is directly maintained.

Those are real-life experiences that we have seen with the use of autonomous networks.

JEANETTE CLARK: I just want to understand that. In your example of the prison setup, where facial recognition allows access to certain areas and not to others, what is the difference between a network that we have known in the past and the self-awareness that you’ve mentioned that you now find in an autonomous network?

SHAMITH MAHARAJ: That’s a great question, Jeanette. Jeanette, I think in the past when we looked at automated processes and automated systems, they seemed to be very different from an autonomous network.

The automated networks give us the opportunity – they are very human-driven – in terms of making the decisions and policing a specific outcome or a specific result.

An autonomous network starts to learn user patterns, user behaviour; what is allowed in certain areas. Using the prison example again, this gives us the ability to ensure that movement is correctly controlled. It’s less resource-intensive, and it’s safer for the employees working in these high-pressure environments.

JEANETTE CLARK: You’ve just mentioned a couple of benefits that make it clear to me why businesses should consider autonomous networks. It’s more efficient, it uses less resources. Are there any others that you can add on, some of the considerations that make this a no-brainer for businesses that should go in this direction?

SHAMITH MAHARAJ: Yes, I think from a cost perspective businesses are under pressure and looking at how to deliver more value in a highly commoditised market. Now this is definitely a cost-saving enabler as well. It’s less resource-intensive and it provides a more application-aware approach to business and to industry at large.

JEANETTE CLARK: But is Africa ready to adopt these networks? Do we have, say, the base systems in place to be able to use it to full benefit?

SHAMITH MAHARAJ: Yes, absolutely, Jeanette. Africa is open for business, and we’ve got the networking building blocks in place with good-quality cable systems on both the east and west coasts of the continent. We have major cloud providers established on the continent, good local content and peering in place.

And more importantly, Jeanette, what we have is talented engineering teams who are highly skilled in artificial intelligence and machine-learning technologies.

More importantly, the customer demand for more application-aware approaches: [these] are available, and customers are requesting more value as well. We believe that autonomous networks is the future, and investing in this today will definitely yield great benefits downstream.

JEANETTE CLARK: Where are you seeing the demand coming from? Are there certain industries that are understanding the benefits they can gain from AI and machine learning faster than others? As a service provider in this area do you see demand in certain areas being higher than others?

SHAMITH MAHARAJ: Yes, absolutely. Jeanette, we are seeing a high adoption rate in the manufacturing sector, and also growing demand in the mining sector. Those would be two of the leading ‘verticals’ or sectors that we are seeing major growth in currently.

JEANETTE CLARK: Can businesses go about starting the journey of autonomous networks with the right IT service provider?

SHAMITH MAHARAJ: I think the most important thing is selecting the right partner. This is of paramount importance and key to success in deploying an effective autonomous network.

Now the process is [that] one will need a trusted partner who is willing to invest the time to understand the customer’s business, identify areas of the business that would realise maximum benefit, and then analyse the big data that’s available, using key tool sets before creating a design blueprint based on the customer and network-behaviour patterns that we would like to see.

To round this [off], you are making use of the AI and ML [machine learning] capabilities in the network in a very controlled manner. Once this is in place, these would be the fundamental building blocks as part of the journey to get a fully autonomous network activated within a customer environment.

JEANETTE CLARK: How long would that process take? It obviously depends on having the right service provider, but putting all of those building blocks in place until a customer can actually have a workable solution, how long would that take?

SHAMITH MAHARAJ: Jeanette, it depends on the business size as well, and the requirements a customer would like to see fully automated, and see benefit from [in] an autonomous network.

From our experience, it can range between three months and a year if it’s a very complex environment. However, the standard that we are looking at currently is approximately three months.

JEANETTE CLARK: I guess the complexity also depends on the legacy systems that are already, for example, in that manufacturing business that approaches you for an autonomous network setup.

SHAMITH MAHARAJ: That’s correct. The approach that we take is always an agile approach in terms of understanding what’s in the environment and how we refresh and adapt the networks to be more autonomous and to provide the value the customer needs. So it’s a very agile approach. From our experience it can range up to a year, but the current average is approximately three months to get the entire environment ready.

JEANETTE CLARK: Shamith, thank you very much.

No matter the application, it does seem as if autonomous networks can open up various opportunities on the African continent, as we’ve just heard from Shamith Maharaj, executive of data networks at BCX, [who discusses] the next step in networking evolution towards true autonomy.

Brought to you by BCX.

Moneyweb does not endorse any product or service being advertised in sponsored articles on our platform.