Petro.ai: Load your data into an AI platform and go. AI is going to tell you things you don’t know. You can’t know without the pattern recognition and tough multivariate calculations of AI.
Petro.ai is the Leader: Then you’re going to push ahead because you’re analyzing, optimizing, and predicting at the frontlines of O&G in the platform that relates science and ROI.
AI in Petro.ai: Hand a calculation-rich machine learning system to your engineer and all those subsurface unknowns become part of models that form option opportunities.
The race is on in O&G. The winners will be those companies that can leverage their data in Industry Intelligent platforms like Petro.ai. There are three important concepts about AI that must be understood to use this model-development space:
1. O&G is falling behind in AI
2. AI is augmented intelligence
3. AI never works alone.
Concept 1: O&G is behind in AI
Dr. Derek Ruths, Chief Data Scientist of Petro.ai explains, “When we talk about oil and gas being behind in AI, what we’re really saying is that they’re an industry that has a ton of data. There are some industries that are not in that place. City planning, for instance, is an area where it would be great to have AI to do city planning, but it’s really difficult to get that data. It’s never been a part of city planning to get all the data that’s involved. Another example is self-driving cars. One of the reasons that self-driving cars have been so hard to build is that we don’t generate that data when we drive. Building an AI in an industry where there’s no data is super hard, it’s impossible—there’s nothing to train the AI on.
“O&G has massive amounts of data. It has great annotations. It has everything it needs. O&G is ripe for AI. So, when we say O&G is behind, we’re really saying there are missed opportunities because they have all the data needed to leverage great AI.”
Concept 2: AI is augmented intelligence
“Think about Fantasy Football,” Dr. Troy Ruths, CEO of Petro.ai begins, “The challenge in Fantasy Football is to create the best team based on the players that are out there and their current performance. AI isn’t making a new player. AI is giving you the best lineup from the current players and their performance. What the AI is doing for you, is if you give the AI your current performance, it can come back with your best lineup. It’s not telling you to drill a new formation. It’s not telling you to try a new completions strategy. It’s telling you what your best lineup is and how to optimize that. It’s giving you things that you can verify, this is real stuff from the real world. It’s just shuffling the deck and giving you back the best ones.”
Concept 3: AI never works alone
Dr Troy Ruths continues, “We’re reaching this point where is it software or is it AI?The line is blurring.You pick up one little thing and everything else pulls with it. That’s the way I think about what Petro.ai is. It’s all these network connections between these different models that are fed by many varying inputs that are all validated. And you can reach into the brain and tweak it a little. Then you step back out and it gives you new results.
“We’ve taken a long time to understand and pull apart the models that need to be geo-based or models that need to be petroleum based or models that need to be data based. It’s an emergent behavior. You put all these models in Petro.ai and what comes out is the AI.
“So, when you’re in Petro.ai doing your job, you’re training the AI at the same time. Nothing happens without context; every piece of the AI is driven by data and by physical principles.Results are verifiable.And you can interrogate it.
“AI needs to be secure. It needs to understand geology. What does Petro.ai need to do? It needs to understand rocks. It needs to understand the well life cycle. And allow you to work in a living system that grows and learns and expands your understanding.”