Robotics: Technology, AI and Automation

 

Robotics: Technology, AI and Automation –

Hi, I'm Piyush from www.swifsys.com, I'm here following the discussion of ai automation. There's a lot of talk about automation in CSPs currently do you think, automation is the is the possible CSP silver bullet for many of their operational challenges.

In this article I am going to explain about AI Automation. I think it is, actually, it's not the silver bullet, I think we're probably in trouble as an industry collectively. I mean, automation is the only way that we're going to be able to cope with that complexity that we've been speaking about already, you know, we've already kind of past the tipping point where networks are just so complicated now that humans on their own can't operate them. So, I think AI Automation definitely is the key, and it's the way that we're going to be able to drive down costs and increase our agility, but it's not as simple as just plugging in some software, switching it on and walking away, we need to do it in a formal manner, and maybe in a stepwise manner as well so we're not talking about Big Bang, it's about doing it in an incremental manner. And how do you see CSPs improving their operational efficiency, I guess, most importantly actually optimizing the use of the people that they have. So I think if you look at efficiency you probably need to look at it holistically so it's looking at technology it's looking at processes, it's looking at people. People are very often, you know, one of the larger components of the operating expenditure for the operator, if not the biggest.

It's about, very simply, making sure that you have the right people in the right places doing the right tasks, and also making sure that automation and AI are augmenting or, or assisting those people so if you've got engineers with 10 or 20 years worth of experience, you need to make sure that they are actually working on the value adding tasks rather than the routine, repetitive actions and that's really where automation can play a key role. So simple alarm handling data build those kinds of activities. If we can automate those processes, it's going to free up those experienced engineers to work on more value adding activities so maybe proactive maintenance may be looking at, you know the needs of enterprise and industry 4.0 customers, and applying some of this telco know how to new 5g use cases, you mentioned AI, how do you see AI and operations, sort of, allowing the future evolution of of operations. 

Well, I think both of those are really critical to the future of operations. And I think you can't actually separate them. I mean, artificial intelligence is going to help you manage this complexity, it's going to bring insights and it's going to help simplify things, but on its own. It's not an awful lot of use. If we want to act in digital time, and we want to have our operations flowing in digital time, we need AI automation to actually implement whatever actions are necessary in the network.

So how do you see us leveraging some of these new technologies to reduce operational costs, where were they, where are the costs that are possible to remove with these new technologies, well there's plenty of scope, across CSP operations to reduce costs, and again I think if you look at it holistically across the whole end to end lifecycle energy would be one example, if you look at, you know the costs of running a network could be half of the OPEX is related to energy costs, you have things like self -organizing networks where you can enable energy saving features and functionality you can obviously modernize your energy supplies. If you look at things like field maintenance, there's a lot of costs, sending engineers out to assign, and if you can use artificial intelligence, you can actually optimize that so they're only going when it's necessary and when you can schedule those visits in a more efficient manner. What do you see as some of the barriers to CSPs being able to adopt AI and automation and what and how do you think we and Naki in particular can help overcome them.

I think there's probably two main barriers I mean it's people and it's data. I mean, humans, potentially are a barrier to anything new from a technology perspective, I mean that's the nature of humans, Sometimes we're resistant to change, and we can overcome that by making sure that we communicate effectively about the end goals of ai automation involving those people in the process right from the start, when you're developing and designing and operationalizing a use case, and also making sure that they have the right skills as well, I mean I think that's a role of employers, and then from the data side, I think we need to realize that there's really no short circuit to this, there's no way that you can take a shortcut.

 I mean you have to just accept that, getting the data is going to take a substantial amount of time, and that's where it's essential as well to have a good mix of data science and telco domain expertise to understand what data sources are available, how to get that data, how to collect it, how to store it, and to make sure that you get the quality from the data because if you think of, you know, a mobile network as an example with 10 million subscribers. You could have 100 billion data points, every single day, hundreds of terabytes of data. What data do you actually want to use, of course it depends on the particular use case, but probably you're interested in anomalies bikes drops, where the system is crossing a particular threshold. It's not even 1% of that data, it's a fraction of that that you're interested in. Without that telco domain knowledge, it's very difficult to know where to get that data.

For more amazing article visit- TechJournal

For more details on what "material is used in 3d printing" visit- The TechJournal.