Energy Industry Insights: Working better and faster with data

VO: This audio feature is part of DNV’s 14th annual Energy Industry Insights study. It is produced by FT Longitude in partnership with DNV.

Steve Edwards:

From value creation to decreased carbon emissions, there are so many reasons why digitalization is critical to the future of oil and gas. In fact, new DNV research shows that 73 percent of energy industry professionals believe advanced use of data is changing the way they work at their organisations. 

So, what kinds of changes are taking place? And how will tomorrow's energy industry participants use data-driven processes to meet the biggest challenges of our times?

I'm Steve Edwards, Senior Editor at FT Longitude. In this special episode of the DNV Talks Energy podcast, we're taking a close-up look at the opportunities and challenges facing oil and gas companies on the road to digitalization. 

Joining me to discuss this is Paula Doyle, Chief Digital Officer at Aker BP. Paula, thanks for joining me.

Paula Doyle:

My pleasure, Steve.

Steve Edwards:

Paula, it would be great to begin by getting a picture of what your role entails at Aker BP and just the pathway you've been on in data and digitalization in recent years.

Paula Doyle:

So, Aker BP, we are a pure play oil and gas company, so that meaning that the energy we deliver is hydrocarbon based, about 80 percent oil, 20 percent gas. We're based exclusively in Norway so the assets we operate are on the NCS, the Norwegian continental shelf. And we are one of Europe's largest independent oil and gas companies.

So, my role there is really around making sure that we're succeeding with digital, working with strategic digital partners, building up the digital strategy and minding some of the key initiatives that we have going on. 

Steve Edwards::

So much change driven by, I think, these twin megatrends of decarbonization and digitalization, and the way that they intersect is really interesting. You mentioned some of your current initiatives. Can you tell us a little bit about the priorities at the moment?

Paula Doyle:

Our strategy as a pure-play player is very clear. So, it's low cost and low carbon. So, as you said, decarbonization is very much key for us. And, as of today, we are one of, if not the lowest, global producer in terms of CO₂ intensity per barrel. And this is something that we think is incredibly important and a real, basically, a licence to operate, but also the type of oil and gas that we should be delivering into the energy mix needs to be as low carbon as possible. And then of course also a low-cost strategy too. So, we are among the top 10 percent in terms of cost per barrel.

And for both of these elements of the strategy, decarbonization and low cost, we believe very firmly that digitalization is the key to unlock it. So, we have been investing heavily in digital, really building up digital competence in the organisation and really getting out there into our operations using digital to transform the way we work and the decisions that we make. This has been a huge part of Aker BP's operations since 2015, 2016, and we're continuing on the journey.

Steve Edwards:

What are some examples of ways in which the organization's becoming, perhaps, more data-driven or even some of the specific technologies that have made a big impact?

Paula Doyle:

Yeah, so one of the major challenges in our industry, and many industries, is that access to data can be quite difficult. So typically, we have a lot of legacy systems. They're locked in. So, you have the data is locked into the application itself. And what we want to do is make data accessible to both humans and machines across our organisation so that we can unlock greater value. Because when people and machines have access to more data, and that data is contextualised, then the decisions that they make will be better and also faster, which is very important.

So, I think one of the interesting things that we have been doing is, for the last eight years, we've been working with a Norwegian industrial software company called Cognite, where basically their mission is to liberate and contextualise data from industrial systems so that it's available to the organisations who need to consume it. And what that means is that, basically, we get the data unlocked from the legacy systems and we get it combined in a central layer so then we can serve it up to various applications.

One of the applications that we serve it up to is our offshore technicians where they have a handheld device. And on that handheld device they have an app which allows them to see all of the information we have in our systems relevant for specific equipment or work orders that they're doing. So, what this does is it really reduces the time it takes for them to get the information, but also enables them to have the information available on the fly when they're doing their work.

Steve Edwards:

When we look at what the leading barriers to digitalization are, resistance to change is the leading barrier to further digitalization. But when you look at a leader group of those who say that their organisation is a leader in digitalization, the top barrier is actually cybersecurity, and resistance to change is lower, which I suppose it makes a degree of sense. Is it your view that organisations that are a bit further along in digitalization have a better appreciation of the risks of cybersecurity?

Paula Doyle:

When we started working with GenAI last year, what we did was put together a cross-functional team. So, we had cybersecurity experts from our organisation embedded with the data scientists, embedded with the infrastructure people because we wanted to get the OpenAI services enabled on our Aker BP tenant so we could create our own chatbot. And within two weeks we were able to work together to get all of the approvals in place for cybersecurity also. So, I'm not sure if it's more of an actual risk or if it's a execution challenge that people have. And here I think cross-functional teams embedding cybersecurity in the execution, getting them involved early in architecture reviews, et cetera, is really key. So, working hand-in-hand with Cybersec for us has been a way that we can keep speed while also innovating.

Steve Edwards:

And is there a lot more value to be harnessed from data-driven approaches?

Paula Doyle:

Absolutely, without a doubt. What we know today is that some of the data that we're basing decisions on is not quality, and also the time it takes us to make decisions is too long. So, for us, it's extremely clear that there's a lot of value for us to be had in being able to be more data-driven. We'll be able to make decisions better, and we'll be able to make decisions faster. And I think it's really important for us as well to be able to use machines where machines are good and use our people where people are good.

Steve Edwards:

What kinds of innovations or technologies are helping with that point you make around the timing, I suppose? Making decisions in real time, getting more up-to-date data, how is that changing?

Paula Doyle:

I think a lot of the benefit that we're getting now and a lot of the value that we're able to unlock is due to our potential or the actual reality of us having applications on Cloud. It allows us to reimagine workflows in a totally different way than we could when we had these applications on-prem. So, it's giving us flexibility and freedom.

I think also the way that we procure software today where we're doing more and more software as a service procurement rather than this large, five-year application licence-based procurement also gives us really good flexibility. So, I think there's a lot of exciting things that have been happening now over the past decade in industry that have moved quite slowly. But actually, when you look and look at the delta from 10 years ago to now, it's quite dramatic in terms of what's available as a digital infrastructure and applications.

Steve Edwards:

Has there been any challenge in getting your workforce to trust data points that you can provide for them? On the one hand, I mean, I suppose there are legitimate doubts about the reliability of some data and margins of error and that kind of thing. But, yeah, on the other hand, you can present accurate data to people who might not act on it if they don't truly believe it's telling them the truth.

Paula Doyle:

Trust in data is key, particularly in our industry which is quite rightfully very risk averse when it comes to operations. So obviously life and environment are number one and people don't want to, and should not, make decisions, operational decisions, based on bad data or hallucinating AIs for that matter. And to be perfectly frank, on this journey, we have done a lot of errors in terms of our data delivery mechanisms. So, taking data from the source systems themselves and up through a stack to present to the user, there we have really been tightening up over the past while.

So, what we have done is we have asked our end users, which is typically our operations and people out in the field, what are their requirements for the data? So, what kind of latency do they need? What kind of quality do they need? What kind of monitoring do they need to see on the data itself through this data delivery stack? And then we work to build pipelines and integrations and the technical delivery to match those requirements. And, for us, this has very much been a journey, and I wish that we had had a better collaboration earlier on, but this is where we are now. 

Steve Edwards:

Paula, what organisational attributes do you think are most important for oil and gas companies that have ambitions to become industry leaders in digitalization?

Paula Doyle::

I think one of them is the entrepreneurial spirit. So, since its inception, Aker BP has been a very entrepreneurial company and I think that really helps the organisation to be able to take more risks, to be able to fail fast and to be able to move quickly. And all of these aspects are really key.

I think one of the main challenges in our industry is that it's conservative, and for a reason, I would say, and also quite traditional. So, the question is to organisations, how do you manage to move in a bimodal setting where you're managing the risk in the way you absolutely need to for your operations, while at the same time you're able to safely and securely innovate on a different speed, on a different risk acceptance and on a different failure level? I think this is probably the biggest challenge for organisations to overcome.

And this is something that we have been working very, very hard with because, typically in our industry, failing is not really an option, right? So, you're trying to bring in this startup software development thinking, fail fast and often, agile approach into an industry where we really focus on de-risking, on being as stable as possible, on making sure that there is no threat to our operations, to life or the environment. So, you have to balance these two sides of the coin and do it in a way that's safe and secure, but also allowing you to have speed and fast innovation cycles.

And I think the challenge, I think, for the industry today is that the innovation cycles are just getting faster and faster. So, we need to speed up more and more. So, if an organisation doesn't fix this challenge, then the gap just gets bigger and bigger and bigger and they get left behind. 

Steve Edwards:

That's really interesting. And I think, so part of the way in which more steps are going to be taken in future is clearly to do with more investments in AI, and there's been a big leap forward. There's been hype, but also a genuine leap forward in recent years. How do you expect AI to change oil and gas operations in the years to come?

Paula Doyle:

So some of the things we're doing with GenAI is we're using it actually to support our exploration teams to identify prospects, and that's a really cool case. So we're giving the model access to many different source systems, including structured data, and it's supporting the process for the exploration teams.

Another thing that we're doing is we're using it to automate parts of our supply chain. So, basically, we're feeding it with documents and we're using it to populate our data model. And our data model, of course, is the core of our digital representation. We found that we've been able to... We populate around 67 percent of the fields with it without fine-tuning, basically out of the box. So, this is very exciting for us.

So, I would say it's still early days in terms of the whole potential area. But, of course, what's extremely important for us as well is to manage the uncertainty in generative AI models, so to keep the hallucinations to zero, particularly in operational settings. As we talked about trusting data, we also need to trust the AI that we're going to use. But there's many applications of it that are not frontline field workers. So, there's many places where we can test it out and we can really minimise the risk. But it's an absolutely exciting technology. In my whole life, I've never seen a technology move so fast, generally speaking, in terms of technological development, but also so fast from consumer to industry. 

Steve Edwards:

Exciting. Really exciting.

Paula Doyle:

Yeah, it truly is. But also, you're back to the data problem because we have to, and this is where we've been quite fortunate with the investments we've been making in data in the last years has come to fruition because if you feed the AI bad data, you get bad output. So really, you're still back into the core of data quality. But what's very exciting for us though is we see, as in the case with the supply chain case, we see the opportunity to use GenAI to populate the data to serve more GenAI. So, it's kind of symbiotic in a sense, and we can see the reduction of years and years of manual effort being able to be executed by code, and that's pretty cool.

Steve Edwards:

Paula Doyle, thank you so much for your time today.

Paula Doyle:

My pleasure, Steve. Thanks for having me.

VO: This audio feature is part of DNV’s 14th annual Energy Industry Insights study. It is produced by FT Longitude in partnership with DNV.