Artificial Intelligence, Machine Learning and IoT are all tools that are barreling through the oil and gas industry right now with a vengeance. This industry is an unassuming recipient of such advances, however they are not only embracing, but growing exponentially due to the safety, convenience and cost driven benefits they offer. Originally a world of pen and paper, rough necks, drafters, engineers, geologists, you name it. They are all being positively affected by the emergence of such technology and it’s taking the oil and gas industry by storm. So get your umbrella, there are many options out there, lots of terminology and some common goals – to drive down costs, improve safety and project forward into the future.
The most important factor in purchasing software is to be educated about it. Some of the technological terms can run together and cross wires occasionally, so it’s vital to know their function and the benefits they offer. Clarification is necessary, because the terms are many times used interchangeably and can create confusion.
- Artificial Intelligence (AI): the theory and development of computer systems able to perform tasks that normally require human intelligence, such as visual perception, speech recognition, decision-making, and translation between languages.
- Machine Learning (ML): the scientific study of algorithms and statistical models that computer systems use to perform a specific task without using explicit instructions, relying on patterns and inference instead. It is seen as a subset of artificial intelligence.
- IoT: a system of interrelated computing devices, mechanical and digital machines, objects, animals or people that are provided with unique identifiers (UIDs) and the ability to transfer data over a network without requiring human-to-human or human-to-computer interaction
Knowledge is power and time is money. They are age old sayings, but they’re inevitably true. The human brain can only understand so much. AI and ML can bring unimaginable speed to processing, streaming, analyzing and reading of big data that is produced by the oil and gas industry. Problems can be solved by experts in remote locations and sometimes, future problems can be predicted and prepared for before they ever happen – so they never happen. It’s called predictive analysis and it’s incredible. It gives companies an advantage like no other. Knowing what is coming and having the ability to prepare for it so there is no lost time.
2Predict is one of the data science services companies out there that specializes in these exact technologies. Headquartered out of California, they provide custom predictive models that deliver advanced insights and data gap analysis commanding highly valued, data driven decisions. They have a team of experts in ML, Deep Learning and AI algorithmic and model development. This also includes the use of Python open source libraries and creating custom algorithms to address specific needs. Their experience is energy specific that spans digital security, high performance computing, networking, supply chain, marketing, oil and gas, and transportation.
I spoke with 2Predict’s Lead Data Scientist, Cedric Fraces, about the integration of AI and ML in the oil and gas industry. “Today AI is gaining traction in a wide range of applications – helping companies streamline processes, improve operations, better serve customers and save money. While traditional Machine Learning still heavily relies on feature engineering, the promise of Deep Learning is to remove human bias from decision making. This dichotomy between big data and expert judgement is a central element of our mission. The reach of AI solutions is still very dependent on industries. Some are ready for advanced deployments while others still struggle with automation and digitization. We help professionals navigate that transition while adding value,” Fraces says.
AI, ML and IoT cannot be set up instantly. Companies have to provide data, implement processes, refine algorithms, etc. The tools have the capability to be smart, but they must go through a certain progression to get to that point. Think of it as an engineer. An engineer starts off as an 18 year old kid going to college for the first time, four years later (possibly much more depending on how advanced the sought after degree) they are equipped with the knowledge to perform engineering functions. That is the process of the human brain. Machines can think and be equipped with so much more knowledge and it can be done much more quickly.
Using 2Predict as an example, there are five phases of their processes:
- Requirements Gathering: Collect requirements and gain understanding of the data landscape, then agree on a statement of work.
- Consultation and Planning: Define the implementation and data scrubbing needs, as well as the algorithms in need of refinement.
- Implementation and Development: Design, architect and deliver the data analytics infrastructure along with any necessary tools, technologies and algorithms to provide the desired outcomes.
- Model Refinement: Customize and refine the infrastructure leveraging ML, deep learning and NLP techniques, and modify the data and processing models until new and unexpected outcomes are achieved.
- Knowledge Handoff & Education: Setup the development environment, walk through some of the algorithms and models and teach how to play with the inputs and observe the outcomes. This level of knowledge handoff may require special skills, but it can be customized based on the skills available to the specific team.
Although there is some leg work involved, it’s only in place that way to optimize performance and maximize the potential of every aspect involved from EPC (engineering, procurement, construction) to exploration and production.
There is another smart technology company that has been in the spotlight recently (no pun intended) as OTC’s recipient of the 2019 “Spotlight on Technology Award.” FutureOn was the only digital solutions provider to win this award for their cloud-based, collaboration application, “FieldAP.”
FieldAP was developed by FutureOn and it is hosted by their own cloud-based data and application platform, FieldTwin. Field Twin is essentially like the app store for iPhone. Major EPC and oil companies are creating applications that will either plug into FieldTwin or be hosted by FieldTwin. Almost like a kind of app store within oil and gas, especially subsea.
“We didn’t want to reinvent the wheel by creating a new engineering software. We didn’t want to reinvent the wheel by creating a new planning tool. What we wanted was to create a wheel hub and spoke model where we were the hub that created the diversity and the technical capability to integrate with a lot of preexisting products or create opportunities for companies to develop new applications that met growing demands or unique needs. What we have created is essentially that. It combines data visualization and integration collaboration, because it’s a cloud based platform, it allows for multiple simultaneous users, as well as global users,” says Thornton Brewer, FutureOn Marking Manager.
The information is a seamless transfer because all of it stays in the cloud-based platform. There are clients with teams in the UK that can work on things and hand off the project to the Houston team as they’re going into work the next morning. FutureOn shifted from a data rendering company to being a data visualization and collaboration company. A lot of engineering data languishes in silos. It’s not easy to get to and difficult to interpret unless you’re an expert. You might have a really crucial well path technology, but if you can’t move that data into a more general population standard format, that data will be pretty much useless to anyone but the well engineer who understands that information. They may not be able to put it in a product or some kind of template that is relatable to all personnel that are working from the seabed up. It’s crucial that all of that information be more readily accessible.
Many times handoff is a major issue. Big companies would (and still do) send large physical boxes of completed documentation in engineering and their clients would have to try to figure out how to integrate that with their filing systems. Down the road, once the field is in operation, the client may go back in search of information, but they wouldn’t be able to go back and necessarily find that document specifically and if they do, they wouldn’t know if it was the most recent.
This is not the case with FieldAP. With minimal training and downtime only being a couple of hours, massive amounts of IoT tech data can be recorded and it becomes usable information in real time, which can create trends and troubleshoot, all at the same time. Companies can make more informed decisions with the right data in a much more comprehensive and time appreciative manner. “We are envisioning this whole life cycle as being fully digitized where no data is ever lost and everyone from operators or decommissioners can see the history of the field. The same way medical records travel with you, they don’t stay at the hospital, they go with the patient. We want this data repository and collaborative eco system to follow the field so that the field has all of this information and operators can use it to create imitative design,” Brewer says.
In the case of the Internet of Things, the “things” that make it up can be any smart device or machine that uses sensors to collect data from the environment that can measure observable occurrences or changes. The data must be communicated to a base system, such as a computer or another device and it must have its own IP address creating a unique identifiable presence on the internet. Most of these items are capable of functioning without physical user interaction.
In the world of oil, it is absolutely vital for equipment and activities to be monitored in every way and for the resulting data, obtained by IoT devices, to communicate with parent companies. The resulting data integrating with AI and ML, can create, document and predict results that the human brain could not. This equips companies with all the possible information that they could conceivably need and provides them with the utmost advantage in a highly competitive industry.
There have been books written about it and movies made about it – it’s the superpower that everyone would love to have: the ability to predict the future. Ironically, those are two of the words contained in the names of the companies featured in this very article. It’s a well-played choice of words, but it is so much more than that. The historical and transactional data that can be used to capture relationships and identify risks and opportunities for the future is a game changer. The technology of AI, ML, and IoT is incredible and until recently, thought to be a purely fictional concept. It is real and the oil and gas industry is not only adopting it, but reaping the benefits. Let me be the first to welcome you to the future!
Oil and gas companies are regularly faced with many industry-specific issues to overcome. Such issues, including exploration and drilling, are often complex and intricate processes with many unique challenges to overcome. Data analytics can play a massive part in streamlining some of the most fundamental operations that are involved in the oil and gas industry.