How Predictive Analytics is Changing Mining and Energy
Written by Simon Meyer
Energy, Metals & Mining
Predictive analytics has revolutionised the process of data collation to calculate patterns and trends. Most commonly used in the retail, financial services, banking and marketing industries, it is now beginning to make waves in mining and energy.
As human beings, our ability to make predictions about the future has been one of the key factors in our survival. Where and when to grow crops, how to achieve success through cooperation, what makes for a sound investment – these are all examples of using the power of prediction to understand the future. Now, this same process is being applied in a much more rigorous and effective way to an ever-increasing number of domains. Be it estimating your arrival time through Google maps, or speculating the future shopping habits of customers based on their loyalty program purchases. Predictive analytics has revolutionised the process of data collation in order to calculate patterns and gauge future outcomes and trends. Most commonly used in the retail, financial services, banking and marketing industries, it is now beginning to make waves in mining and energy.
The status of predictive analysis today
According to States Lloyd’s Report from 2018 , oil and gas companies are slowly starting to develop their own predictive technologies. For example, between 2012-2018, 5 out of 425 worldwide patents filed in predictive analytics were related to oil and gas operations. Additionally, mining and energy are laborious, equipment heavy industries, that are subject to unexpected disruptions, such as equipment failures. This can have costly knock on effects on production times.
In recent decades, mining in particular has been the victim of a production downturn. Although miners were receiving a lot of data from the sensors used in their underground operations, only a small percentage was in fact accessible. Not being familiar with using IT as a strategic tool was perhaps a contributing factor.
However, despite a recent turnaround in the mining industry, key players are recognising the need to improve efficiencies within production. Some of the top global mining and energy companies such as Shell, Chevron and Schlumberger have turned to digitalisation in search of a solution. In order to maximise production, lower costs and improve safety, they are utilising predictive analytics in a number of areas including transport route optimisation, oil and gas trading, and demand forecasting.
Practical use cases of predictive analytics
The latter of these is especially interesting to Schlumberger, as their CIO Eric Abecassis explains. “We’re using predictive analytics in demand forecasting to manage our inputs and resources more efficiently. We need it, for example, to allocate tools and equipment to projects across 85 countries and to plan raw material inputs. We also use it to ensure people with the right proficiencies are available when projects get under way”. It’s easy to see the benefits this can bring to companies, given how capital intensive tools and equipments can be.
Another area where predictive analytics technology proves invaluable is in the maintenance of equipment. Embedded sensors and remote connectivity allows for essential monitoring of vital production processes such as temperature, vibration and pressure. This results in issues being resolved proactively and avoiding any major failures or delays. It also removes the requirement for manual inspections, which has resulted in a safer working environment. The ability to provide detailed predictive analytics solutions through real-time dashboards, alerts and equipment prognoses is helping forward thinking mining and energy companies increase output and reduce downtime.
Margery Connor, Chevron’s leader of data science capability and technology, outlines the direction this will take the company in. “For us, the next big predictive analytics nut to crack is maintenance,” she reveals. “We are making progress in preventive maintenance and in improving asset integrity. We will use it to predict events and better plan maintenance and inspection schedules. This can be applied in upstream, midstream and downstream”.
Predictive analytics as an enabler of the future
This prediction of future events and helping to determine risk assessments and work-related scenarios has proven a valuable tool in the continued advancement of mining and energy. But it seems the industry is keen to always be pushing further ahead. In fact, energy companies are seeking even more streamlined ways to harness global data. Despite significant advances, the framework of predictive analytics is not always clearly defined. Changing variables can result in inaccurate analytics and thus the potential of this big data tool being wasted.
In response to this, Axora offers predictive analytics solutions on the platform that seek to interpret the extensive data a business collates and extract the vital insights that really impact a business and its progression. Through advanced AI and machine learning algorithms, they can predict, for example, the most important behaviours of customers, such as their loyalty to a business. Or, they can provide a more efficient reporting of data when taken from multiple global sources.
Companies using predictive analysis are already seeing beneficial impacts. As an example, according to the States Lloyd’s Register report, “Occidental Petroleum has quantified the savings generated by the use of machine learning to predict drill-bit location and identify optimal bit-turn rate at $325,000 per rig TransCanada, meanwhile, estimates that in 2016 it avoided costs of $7 million thanks to its analytics platform’s success in predicting failure of its gas pipeline assets in the eastern US”.
In addition, with increased talk of cross over between industries to further advance the potential of mining and energy, there’s a myriad of ways this could develop. And with more companies recognising the benefits of these data tools to their business and processes, we have identified a number of relevant assets in the predictive analytics space.
Get familiar with our recommended solutions!