Abstract

Global oil demand is at an all-time high for industries, transportation, technology, and is forecasted to increase by more than a third by 2035. Approximate 95 million b/d of oil the world now consumes, up from 86 million b/d in 2008 and an 11% rise even amid the worst economic times since the 1930s. With this surge in energy demand, the oil and gas (O&G) industry is required to improve production efficiencies and strong safety culture, i.e. health, safety and the environment remain non-negotiable imperatives.

One offshore platform costs billions of dollars. Any small hazard and an improper decision result a huge monetary loss, human life, resources to a company. Hence, the safety behavior, human performance, and production aspects have been a top priority in the O&G industry. This industry has undertaken extensive actions to enhance process safety and human performance and strictly implement proper decision and emergency control measures. With significant oil price decline recently, O&G companies have responded by utilizing digital transformation technologies to gain better insights, drive operational efficiencies, production improvement and ensure survival.

Abstract

The recent oil price collapse has triggered a wave of cost reduction in Oil and Gas (O&G) companies. As one of the most sensitive and dangerous industries, the infrastructure in O&G companies requires regular monitoring and maintenance to avoid accidents that can lead to environmental problems which result in increased operational costs. Digital technologies are recognized as adding value to O&G companies by helping not only reduce costs, but also increase workforce productivity. Modern O&G companies utilize digital technologies for obtaining information about ongoing operations to gain operational efficiencies, production improvement, and safety performance. As a result, modern rigs often have a number of video-based sensors actively monitoring the exploration and production process. However, most of the video-based monitoring activities have traditionally relied on manual methods as such increase the operational cost.

Computer vision (CV) technologies facilitate advanced video interpretation that allows achieving continuous, robust, and accurate assessment of many different phenomena through real-time and pre-existing video data. Recent advances of CV combined with deep learning (DLCV) can lead to significantly improved performance of monitoring process for improving safety and productivity performances. The main challenge is devising the technology while adhering to the constraints of operational cost and data processing latency.