Computer Vision and the Oil Industry

One of the missions of artificial intelligence is to help people work in hard-to-reach places and hazardous enterprises. The oil industry, which involves working at remote fields, can be called such, and technology like computer vision significantly eases the extraction of this natural resource. Since 2018, there has been a rapid growth in the use of technologies involving computer vision in the oil industry, and it is expected that the need of oil corporations for digitalization will only increase in the coming years. There are several main tasks that computer vision solves in oil extraction, namely: remote control, production safety, and production optimization.

Oil fields are located in remote areas, which are often difficult to reach. This is what causes problems related to the physical control of oil drilling rigs — it is necessary for teams of people to always be present at the station, which often carries risks and costs additional resources. Computer vision helps to solve this problem: in addition to reducing the number of trips to remote fields, it increases the automation of data collection. Computer vision is not just video surveillance systems, where the video stream only provides an image, and a person still extracts information from it; computer vision is also remote monitoring of the field, where computer systems extract data from the video stream, warning the crew of any malfunctions at the oil drilling station.

But what to do with the data that computer vision has extracted and provided to the station employee? Of course, monitor the safety of the enterprise and prevent dangerous situations. An oil drilling station is a dangerous production facility with high temperatures and toxic vapors, so it is especially important to monitor compliance with all safety measures, including the presence of personal protective equipment for employees and control over hazardous areas. Moreover, computer technologies have become so advanced that it is possible to detect leaks at drilling sites. For example, Osprey introduced its own computer program at oil production facilities, which detects critical methane leaks from storage tanks thanks to built-in thermal imagers in surveillance cameras, and this reduced the number of crew trips to the fields by 50%.

Oil rigs are constructed of metal structures that are subject to wear and tear, and an unexpected breakdown can cause an accident. When a person monitors the quality of equipment and structures, there is a high probability that he may miss something or check very slowly. However, thanks to computer vision, control over equipment can take place much faster and more thoroughly, preventing accidents caused by equipment wear or metal corrosion. What is important — timely detection of a defect can reduce production costs several times. For example, V-Soft Digital introduced computer vision to detect defects on steel pipes used to transport oil. In addition to increasing the accuracy of defect detection by 17%, this also allowed to reduce costs on one production line by $150,000.

Given that artificial intelligence solves not only production problems in the oil industry but also economic ones, the rapid growth in the use of technologies with computer vision becomes quite expected.