Picture the scene:
You’re one of the ground crew at San Diego airport.
It’s night-time, it’s baking hot, and you’re coming to the end of a long shift.
Your overalls feel heavier than ever, clogged with sweat and dirt of operating machinery all day.
Your limbs ache from having to lift heavy refuelling hoses.
Even on a good day, and even though you’ve worked at this airport for five years, you still feel slightly nervous for the pilots flying over the mountains.
Tonight, it is lashing down with rain, visibility is poor and there is a wild wind blowing from the west.
At the beginning of the day you were excited about the challenges of the day’s work, about what you would tell your partner later; all you want now is to be home in their arms.
Only two more refuellings to go…
In situations like this, the risk of human error goes off the charts.
What happens if one of those two aircraft that land is completely unfamiliar to the ground crew?
Imagine you have to leaf through field notebooks to assess the decals on the aircraft – the symbols which tell the engineer which exact fuel to put in - and then match the image to the correct fuel?
Imagine getting it wrong, and the aircraft suffers engine failure shortly after take-off.
In situations like this, AI can be incredibly useful.
Five years ago at bp, we launched a tool called safe2go which uses machine learning and computer vision to match the decals to the correct fuel.
The ground crew can use it to scan images – even ones the tool has never seen before.
Since then, it has processed over 5 million refuellings and protected over 800,000 overwing fuellings – which are more dangerous.
And it’s done it with a 100% success rate.
We know all too well how inherently dangerous our industry is.
That’s why we put safety at the top of everything we do in bp – because we care about our people. And because we care, we are working with AI to see where to best use it. Done the right way, AI's potential is huge.
And right now, we need every tool we can get as we face one of humanity’s greatest challenges in climate change.
And that challenge is no longer a scientific challenge – it is an engineering challenge.
Because if the world is to get to net zero, it needs to completely rewire its energy system.
And the shocks to supply we have had in recent years show that we can’t simply switch from one system to another.
People want and need energy which is secure, affordable, and lower carbon – and bp is set up to address all three.
The transition has to be rapid and orderly: the world needs to invest to accelerate the transition, and – not or, we need to invest in today’s primarily oil and gas system to keep energy flowing where it’s needed.
Three years ago, we announced our new purpose to reimagine energy for people and planet and our ambition to get to net zero by 2050 – and to help the world get there too.
Three years ago, we set out our strategy to achieve this.
And in three years, we went from 3% of capex to 30% into what we call our transition growth engines – biofuels, convenience, EV charging, hydrogen, renewables and power.
And as an engineer, I am proud to be working in a company that has the scale, the skills and the experience to help play a part in this.
But we need help to succeed.
We need you.
In 2020, we began laying the foundations for a different bp.
A bp that would think, look, and feel different – and lean into the diverse community of talent to evolve.
As well as being an engineer, I am the chief scientist, and chief digital and technology officer – and I head up the amazing Innovation and Engineering team, which is made up of experts in everything from safety to advanced digitalization.
And last year we set out a new vision of digital in bp.
I am not satisfied for us to compete against other big energy companies.
We are now benchmarking ourselves against the best tech companies.
We’ve changed from a big portfolio to a product-led operating model.
This means we are laser-focused on building solutions to problems faced by our customers and businesses – especially on safety and bringing down emissions.
And we are aiming for 70% of our digital staff to be technical by 2025 – meaning coders, coders, coders!
We need more talent – because we want to build, build, build!
We need more people who are going to bring in new ways of working, bring diverse experiences from other industries, and create new models of innovation that we haven’t even thought of yet!
We need people to help us transform a 114 year old oil company into a nimble, sustainable, and tech-driven energy company.
And a great part of this for me is to lead the energy sector in the application of AI.
For us, it’s about the bringing together the physical and the digital – using one to enhance the other.
To give you one small but vitally important example, we can use AI to monitor corrosion in a way that we just can’t do manually. It means we can spot risk faster – and cut those manual inspections by 70%.
But we need to go further, and we need you.
We’ve been working on AI and machine learning since 2018.
It’s already helping us choose the right locations in the world to build hydrogen plants, the best locations to install the next EV charger.
And within our oil and gas business, we’ve built a tool that I could only dream of when I was heading up our technical operations five years ago.
It allows our biostratigraphers – or the geologists who date layers of rock according to fossils – to reduce the time it takes to understand how old a site is and how it was formed. A process that used to take two months now takes two weeks – and allows us to decide much faster if we should drill in the area.
And the coolest aspect about this tool is the multidisciplinary collaborative effort that went into building it.
Data scientists, geologists, engineers and architects - geeks and non-geeks working together in perfect harmony.
But we need to do even more than this, and we need you.
Imagine industrial virtual reality.
It would build on the digital twins we are already using to digitally map our physical assets.
It would build on the remote collaboration centres we use to help our offshore teams monitor activity from afar.
It would mean that the head of our oil and gas, or offshore wind, or hydrogen businesses could wake up in the morning, see what is happening across all the operations worldwide and optimise processes by simply asking their virtual assistant for a recommendation.
And that’s where we need you to help us get there!
So you can see the potential range of applications of AI across energy is enormous, and it spans several different industries.
Optimisation of processes, efficiency, cost savings, lowering emissions, and unlocking growth – these are a huge part of what we are trying to do at bp. But beyond that, there’s another consequence of AI that I’m excited about.
To talk about it, I’m going to refer to a book.
Despite being a tech geek, I’m still obsessed with printed books.
So much so that not long ago I managed to read 52 in a year.
In fact, I even make room for them in my small suitcase when I’m out on the road!
These are books written by humans – but even if they were written by AI, I’d still read them in print if I could.
This for me is the perfect balance between the digital and the physical – both enhancing each other.
There’s one English author whose writing I’ve always found interesting – and that’s George Orwell. In his 1937 study and essay, 'The Road To Wigan Pier', he wrote about the conditions of coal miners in the north of England.
Orwell was a sceptic of mechanical progress and what he saw as machine worship.
For him, it was not just about machines being using for destructive purposes – and we do really need to be careful that we don’t create more division in the world with AI and algorithms that are biased or which discriminate.
But he poses another important challenge, that since people naturally tend towards effort and tasks, a fully mechanised world is one in which there is nothing for humans to do.
However, I don’t think Orwell envisaged a world as complex as ours now.
And, understandably, I don’t think he foresaw a world in which people were working to rewire the entire energy system.
What are machines good at? They can quickly analyse and find patterns across large volumes of data.
What are people good at? We are great at creativity, intuition, and logical reasoning.
People in bp tell me how excited they are by the potential of AI to remove mundane tasks so they can focus on innovation and the problem-solving.
And so instead of machines replacing humans, at bp we see AI about humans using machines to form human-machine teams.
The solutions we need for the energy transition are going to require the strengths of both people and technology.
The creativity that AI frees up is integral to the process of finding the engineering solutions to make the energy transition a success.
And the opportunity at bp to be part of that success story is huge.
Ultimately, we are investing in AI not just because we want to improve sustainability, efficiency and earnings, but because we want to empower our people to work better too.