View Full Version : Best Practices in Data Science,Lessons for the Oil and Gas Industry

01-21-2021, 06:37 AM
Many oil and gas executives are concerned that their industry is behind in the race to apply advanced data science — including machine learning and artificial intelligence (AI) — to its most critical problems. This is understandable considering the complexity of the sector’s challenges. First, the cost of failure for oil and gas is significant, restricting the amount of experimentation companies can undertake and pushing them to focus on high-quality solutions. Furthermore, many activities, such as developing a field, happen relatively infrequently, making it difficult to obtain data at the required scale to develop algorithms. Finally, data is often extremely valuable, which reduces the incentive for cross-industry sharing. Despite these challenges, when AI is applied to the appropriate areas, the impact is likely to be significant (see figure).

Delivering sustained business value from digital technologies–particularly AI–requires good problem formulation, data readiness, expertise availability and organizational enablement.

Addressing the right problem
Not all problems can or should be addressed using advanced analytics. AI solutions are generally appropriate for two broad classes of problems: 1) complex business decisions that hinge on predictions inferred from historical data patterns, and 2) automation of commercial or operational processes with complex but discernible underlying patterns.

For example, GE determined that it could improve the effectiveness of equipment maintenance by applying predictive algorithms to heat loss. If anomalies in heat loss are handled proactively, operators can avoid unplanned and costly downtime for repairs. However, because certain critical components did not contain sensors, they could not be easily monitored. To address this, GE developed a heat-monitoring smartphone app called TITAN that uses an iPhone equipped with a thermal camera to provide noninvasive monitoring of components and flag instances when equipment needs repair.

Gathering the right data
AI and machine learning algorithms frequently require significant amounts of data. The data must also be of sufficient quality and granularity, and it must be appropriately representative of the phenomena being modeled.

BP was seeking to reduce fugitive emissions, particularly in certain mature fields. To create machine learning algorithms that addressed this problem, the company needed enough data to develop and test appropriate models. However, outfitting all their wells with sensors to gather data would be very costly. BP and its technology partners came up with an inexpensive solution: They fixed Android phones to beam pumps and combined the data gathered with historical maintenance logs and weather recordings. This allowed them to test the algorithms and prove their effectiveness.

Assembling the right expertise
Effective application of AI requires more than analytics expertise to ensure the right AI tools and technology are being implemented. Other capabilities are also critical, including domain experts and individuals who can close the gap between technical skills and commercial understanding.

To help optimize its overall energy portfolio, Exelon wanted to accurately dis***** excess power generated by its wind turbines to the energy market. In order to do this, it needed a five-minute forecasting capability to predict when wind speed would change suddenly. The company was looking for an OEM-agnostic data aggregation and analytics solution, but did not have all the required capabilities and did not want to shoulder the risk and cost of in-house development. It therefore used a partnership approach that ultimately increased annual energy production by around 3%, and reduced operating costs by 25%.

Ensuring the right organizational configuration
Scaling up AI solutions across multiple areas of the organization requires a receptive organization. Senior leadership needs to take ownership of the process to encourage technology adoption. Operations-level “evangelists” can help share lessons learned and expertise across the organization.

Rio Tinto combined its in-house mining and analytics expertise with the specialties of partner companies to develop automation solutions for drilling, extraction and ore transportation. It also created a dedicated data science unit that disseminated ideas across business units, and it promoted an open innovation culture via an innovation lab and ****athons.

Powering up AI solutions
Clearly advanced data science applications have a place in the oil and gas industry, and their potential to yield tangible benefits is considerable. But unlike other industries, the sector faces challenges due to its complexity. Nevertheless, many companies have already developed creative data science solutions. Applying their lessons learned within the right framework will allow AI to be deployed successfully across the industry.