AI and Automation Transform Mexico’s Oil Exploration: Apogee

Q: How would you describe your company’s role in Mexico’s oil and gas industry?

A: We are a Mexican technology services company that seeks to increase hydrocarbon production in the country, and ensure that all our services generate robust exploratory portfolios. In the end, we do not want dry wells. How do we achieve this? We rely on technology, the latest that the market has to offer. We are also currently in the process of developing new technologies. Our goal is putting them at the service of the client to make the most out of information. 

Q: What role do automation and AI play in your services?

A: Processes that were once carried out manually are now automated. Computing power has allowed us to implement things that we already knew could be done, but which before were not feasible. For example, seismic interpretation used to mean that an interpreter manually picked positives and negatives. Today, with an algorithm, you plant a seed, it interprets everything, and the interpreter’s role is to assemble and validate. Then there is AI, which is here to stay. I compare it to when we moved from the typewriter to the computer: there were always people who refused to change, but that did not stop progress. This is the same. All our services today involve automated processes, computing power, and AI. The results are surprising: processes that used to take three months can now be completed in 10 or 15 days, depending on the data. At the end of the day, that means cost reduction, process optimization, and adapting to a changing market. In the past, a full reservoir study could take two years; today, it takes two or three months.

Q: How is oil and gas exploration evolving today?

A: Exploration is currently in the process of breaking paradigms. Why? Because the prime targets, the big onshore fields and the giant structures, have already been discovered. Now, we must move to the frontier, deepwater. Why deepwater? Because it remains unexplored. If you look at Brazil, West Africa, the US Gulf of Mexico, new giant fields have been found. Guyana is another example. Here in Mexico, we also have these large fields waiting to be discovered. But this brings challenges, not only technological but also financial. An onshore well might cost US$2-3 million, while a deepwater well can cost US$100 million. That is the direction we are moving toward. In the past, producing from shales or metamorphic rocks was unthinkable. Today, those paradigms are being broken because we must go to that new frontier. The easy hydrocarbons have already been found.

Q: How do you incorporate global technologies into Mexico’s energy market?

A: In geoscience, we are at the forefront, bringing in technologies that are already proven globally and incorporating them into the Mexican market through an integrated workflow. This allows clients, including not just PEMEX, but also others private companies, to adopt technologies that otherwise would be difficult to implement because of existing platforms and databases. We take these technologies, integrate them into our automated processes, and deliver the final product in the format and conditions that the client requests. How does this help? By accelerating processes. For example, a dynamic model with millions of cells that used to take weeks or months can now be completed in days using GPUs and the cloud. This gives clients more time to analyze data, assess uncertainty, and evaluate risks. Ultimately, this reduces uncertainty and increases success rates.

Q: How is your company expanding its services and global presence?

A: Thanks to the exponential growth we are experiencing, we are now expanding our services. We are evaluating and closing agreements with technologies for implementing AI in production, operations, and drilling processes. At the same time, we are beginning internationalization, with projects already in West Africa and Brazil, using Mexican talent. The pandemic helped break the paradigm that you must be physically at the client’s office to carry out and complete a project. Our goal now is to consolidate our geoscience offering, and next year, expand into new areas of the oil industry.

Q: What role do alliances and partnerships play in your business model?

A: Alliances are a win-win. They help us to access new technologies that put us two or three years ahead of competitors, while giving partners immediate access to the market with almost no investment. They provide us with the technology and representation, we test it, and if it works, we incorporate it immediately into our services. We always avoid competing with partners. If we already have an agreement with a company specializing in automated seismic interpretation, then we look for complementary solutions that expand our services. This also makes it easier for the market to adopt.

Q: Can you share an example of how your technology has accelerated results?

A: At the Mexican Oil Congress, we presented our work in exploration. For example, we interpreted 9,000 km² in less than a year, when traditionally 500 km² would take the same time. In another case, in the Gulf of Mexico, we used an AI algorithm originally developed for processing images and photos in seismic data collection. Interpreting 500 km² of salt took just five or six days. In that short time, you already have the structural base. Then the interpreter adds geological knowledge. What used to take months now takes a few days. This gives managers an early idea of what is happening and what can be found, while leaving time for geological refinement.

Q: How has automation changed seismic interpretation work?

A: In seismic interpretation, an interpreter used to work on four, five, or six horizons, spending most of the time generating them and little on quality control. Now, we can process 100, 200, or 300 horizons in less time, in our case, using Paleoscan. Quality control is more robust, but the overall time is much shorter. Before, four or five people were needed; now, two are enough. Still, the critical task today is ensuring quality and verifying scenarios. At the end of the day, I always tell young people: this is a machine, and thankfully it remains a machine. It does not understand geology, it solves a mathematical problem. We must give it geological meaning. Interpreters and reservoir engineers will always be needed. We cannot deny change; we are moving from the horse-drawn cart to the automobile. This is happening, and it will not stop.

Q: How has the quality of data evolved in the oil industry?

A: The evolution has been tremendous, from how data is acquired to the type and quality of data. If you compare seismic from the 1970s to today, it is like high definition versus something blurry. Everything has evolved hand in hand. Automatic processes now use much better input data, with clearer images and more precise measurements. For example, production used to be measured analog and now it is digital.