Julian Alfred, Sales Enablement Lead- Global Energy Program- Dell Technologies.
So, you're one of the keynote speakers from this morning, what were the key themes of your presentation?
Well, as you know, the oil and gas industry is really trying to optimise the way in which we do production. We want to reduce costs. We want to keep operation safe and one of the ways in which companies are trying to address that is by using a lot of data to understand how their assets are performing, and then to run analytics on that and then to optimise the way in which they run operations. One of the things I was trying to talk to the industry about is how far should we be taking analytics? So, there's no question needs to be done but how far do we want to automate system to the point where the analytics is effectively running the show? Who is responsible in the end when something goes wrong perhaps but ultimately, how are we going to validate the decisions that artificial intelligence is actually making? We're a long way from that problem at the moment with so basically we're focusing on trying to help companies to understand how to gather data from the Edge transmit it to a central environment so we can perform analytics and then create models that we can push back out to the Edge so that we can make the Edge more intelligent, so decisions could be taken where the asset actually is at in a more automated kind of way and that way we think with this sort of distributed intelligence the oil field operations can be more optimised and can be safer.
You had quite an interesting analogy that you made on one of your slides, you had a picture with an octopus?
Yeah, I think that was interesting because an octopus has an essential brain, and then in each of its eight tentacles it's got a mini brain. Now what's fascinating about that is that the tentacles can in some ways make decisions on their own, which kind of is analogous to Edge analytics where we’ve got assets out in a state of North Sea where you want to put intelligence into those assets so that they can perform well, in an optimal way, they can self-heal.
So two things, one is to feed the data back to a central process so that we can understand better how the asset is performing, but then when we push models out to the asset itself, we can help it look at its own data and optimise its own behaviour as well and that I think will create a much more efficient operation. However, I think back to the octopus analogy, is that even though each tentacle has its own mini brain what’s fascinating about an octopus is that there is a central neural ring that connects all eight of these tentacles. So, what happens is that the tentacles can actually collaborate with each other and that's exactly the same kind of thing we want to have with oil field operations. It's not only to optimise assets on their own but to have assets cross-optimise themselves so that overall the field asset itself behaves in a very optimal way.
Going back to what you said earlier about the human factor, you were very clear that wouldn't be replacing this stage in we have to kind of have a look at liability, how far do we take this artificial intelligence and what do we want to doing so that we still have a human in control to make those management decisions which are obviously key from the operator side of things?
Yeah, and is very interesting, I mean, there's no question that the technology that exists today if you want it to fully automate an asset you could do, perhaps over time, as we get more data and we have better algorithms, faster computing, you can have an asset operate completely on its own but is that where we really want to get to? I don't think it's a good idea for us to advocate complete responsibility to artificial intelligence and just simply let it run. I think whenever we build automation into our systems the human must have a place in terms of being able to make an ultimate decision and to at the very basic level to at least understand what the AI is ultimately doing with the asset. To put it another way, whatever happens with automation, we need to be able to understand how the asset is performing, why it's performing that way and if artificial intelligence is making decisions on the behaviour of the asset, we need to understand how those decisions were taken so that if we need to step in we know how to do so in the most optimal manner.
Lastly just to kind of to finalise for those who don't know. How active are Dell Technologies within the energy sector just now? I mean, what's it worth to the company?
Dell Technologies, you know, many people look at Dell and will obviously know it from laptops and monitors, but Dell Technologies is a 90 plus billion dollar organisation and we have technology ranging from laptops and monitors all the way to high-performance computing to help with things like syphon processing all the way through to converge infrastructure to help geoscience applications be provisioned to the geoscience community in a much more optimal way all the way through to scalable data management environments because in the oil and gas industry to talk about there's a lot of data, petabytes of data that needs to be managed so you need to be able to scale easily and Dell Technologies offers a wide range of solutions that enable oil and gas companies to be better in the way that they operate. But specifically, in terms of what we were talking about today with artificial intelligence, IOT is one of the key ingredients that makes up successful connected digital operations oil field and that is what we offer. We offer the Edge Gateway technology, ruggedized and also the ability to run analytics on the Edge and we also offer the core analytical capability to actually drive that big data problem-solving and then face out to the Edge. We do that by collaborating with other partners within the industry that have very specific analytical solutions that we can provision with our IT infrastructure.