Rise of the Machines: How to Leverage Machine Learning and Artificial Intelligence to Optimize Your Oil and Gas Workflows09/29/2016
Proper use of software can save countless man hours, but only if care is taken
This is the first in a three-part series about machine learning's applications in upstream oil and gas. This blog details how machine learning can be used to eliminate tedious tasks, enabling teams to focus more of their time and energy on analysis and optimization.
Have you ever said to yourself, “There has to be a better way”? If you’re like most, this question pops into your mind multiple times a day.
Technical staff and managers at every size oil and gas company use this question to drive innovation. Countless innovations have been spurred on by this simple question. In the early days of the oilfield, innovations like rotary drilling, artificial lift, and advanced formation evaluation techniques were led by men and women who refused to accept the status quo. This simple creed led to safety and technological advances unimaginable by early pioneers.
Computers and software are the tools of innovation in the modern oilfield. In the past, the list of tedious tasks technical staffs were required to complete was never-ending. Daily production, drilling, completion, and other activity reports had to be completed daily. As oilfield software was developed and brought to the forefront, innovators used these new tools to shift man hours from processing tasks to more valuable analytics-driven work.
The goal for all operators is to use computing power to eliminate time-sucking tasks and focus staffs on innovation and problem-solving. Most simple tasks have already been replaced by computers and simple programs, but the next wave of innovation will require something more: machine learning and big data analytics.
Machine learning is the science of designing computers to work without explicit instruction. Think of machine learning in this way: When a staff member uses a program to forecast production, they guide and supervise the computer well by well until every well has a forecast. If machine learning is implemented, a group of wells can be selected and with one click of the mouse, the program will forecast every well without direct supervision. Instead of spending 80% of their time forecasting and 20% of their time analyzing the forecasts, machine learning helps shift man hours from tedious tasks to analytical tasks.
Field staffs are a common casualty of tedious tasks. Field personnel are most efficient when they are able to maximize their time spent interacting with and monitoring wells. But many field workers are forced to spend much of their time in a field office typing up reports and inputting data. Machine learning and big data meshed with SCADA can eliminate these tedious tasks. Programs can be designed to query data from SCADA, build daily reports, and email the reports across the company. This saves valuable man hours and allows for more efficient use of time across the field.
How does machine learning fit into your workflows? Think back to the times you have said to yourself, “There has to be a better way.” Many of the repetitive tasks you dread can be automated. Harnessing software and computing power can remove the tedious tasks from your workflows and allow you to focus on the fun part of your day – analytics and optimization.
Optimize Your Production Operations
Speaking of optimization, check out the resources on our production microsite. This site contains customer-success videos, case-study webinars, and white papers – all designed to help your organization optimize all facets of your production operations.