Q&A: AI’s transformation of the energy industry, with Shell’s Dan Jeavons

Global energy giant Shell has invested heavily in AI and analytics as part of a digital strategy that has saved it over a billion in cost reductions, production increases and additional customer margins. Heading up the company’s data science team is Dan Jeavons, a data expert and digitisation enthusiast who has been repeatedly recognised for his contributions to the field. Ahead of his keynote session at London’s Big Data World next week, we asked Dan about Shell’s hugely successful data strategy and where AI’s benefits are being most keenly felt

When did you first truly appreciate the power of data, analytics, and AI?

I’ve spent my career looking at how to drive business value from data – but it was around 2013 that I really saw the change starting. The new tools and methods coming to the market started to get really exciting. I saw the potential and so pitched for a role in this space. Since then, it’s been an exciting journey.

What are the key ingredients of Shell’s data strategy?

At the end of the day our data strategy is not really a standalone – we are trying to embed data strategy into our business strategy. We are trying to understand which KPIs are going to move the needle for the business and then use that to understand which data is needed, what quality it needs to be, and, of course, which AI opportunities to pursue.

How is Shell incorporating AI into its data strategy and why?

AI is at the core of our digital & data strategies. Most digital technologies create or consume data. AI provides the opportunity to create new insights from data by turning it into information. It also enables us to change our business processes by eliminating manual tasks. We see AI as a key enabler for value from digital technologies and have made an active decision to embed AI in every part of our organisation through the Shell.ai programme.

What sort of AI are you implementing?

We are already using a variety of AI methods – focussing on applications like machine vision for robots and drones supporting use cases like inspection, using machine learning methods to identify potential failures in pieces of equipment, using natural language processing to extract key information for our analysts from vast quantities of text and also leveraging optimisation approaches to provide smarter electrical supply electric vehicles. The list goes on but that gives you a flavour.

For Shell, what are the technology’s most promising use cases?

I think the real promise of AI is not related to one specific use case, but rather to the ability of the implementation of AI at scale to completely disrupt the way in which we do business. If you take the management of our assets – the physical facilities that produce energy, whether wind farms or refineries – AI gives us the opportunity to inspect, maintain, monitor and optimise these facilities in new ways. The role of the operator, the maintenance technician or the inspector will change a lot over time. That’s not to say these roles will disappear, but what they do day to day will be very different.

What are your objectives for 2020 and beyond?

2020 for us is all about scale. We want to roll-out a large number of AI-based applications and have them used consistently across our organisation as part of our standard business processes. We are focussed on applications like predictive maintenance, real-time process optimisation and inspection. We think this is possible and we are well on our way.

What lessons have you learned so far and what have been the results?

My biggest learning is that AI deployment is less about the technology and more about change management. As someone with a technology background, I spent a lot of time focussing on the technology challenges of deploying AI at scale. That is important but the bigger challenge is learning to encourage adoption. The Shell.ai programme is really starting to help us with this.

What is your key message to the senior data leaders attending BDW?

Don’t underestimate the impact AI is going to have in the coming years on your business. This technology is becoming mature and the deployment of it is going to accelerate in the coming years. Adoption will be the key to remaining competitive in many industries. I really do think this technology is going to cause significant disruption because the potential applications are so broad and pervasive.