The AI revolution: How a strong data strategy is giving energy leaders an advantage

March 11, 2024

The AI revolution: How a strong data strategy is giving energy leaders an advantage

March 11, 2024

abstract image representing AI and data

Real-world applications for how a solid data foundation and strategy can transform work in Canada’s energy industry.


Jason Lee is a Partner and MNP Digital’s Applied Data & Analytics Lead. Drawing on two decades of experience in technology, project and account management, and data innovation, Jason solves complex technology and business challenges to help his clients thrive.

The digital revolution happening around the world has reached a point of inflection with the widespread acceptance and use of artificial intelligence (AI).

Across industries, AI plays a central role in driving innovation, efficiency, and sustainability in ways many people only dreamed possible. The energy sector is in an optimal position to build off these advances to increase sustainability and be at the forefront of this movement achieving excellence in narrow AI (where a learning algorithm is designed to perform very specific, assigned tasks).

In fact, we see opportunities abound to leverage the technology available and develop a forward-thinking approach to its use in the future.  

With so many applications, we’ve called out a few impact areas in which the energy industry is adopting AI to help reduce cost and risk in operations, help attain sustainability goals, and generate new opportunities.

Acting as inspiration or as a clear directional goal for your organization, it’s important to see the various ways this technology is being employed to amplify innovation and productivity.

AI uses in exploration and production

Adopting machine learning algorithms in exploration and production is ground-breaking. These algorithms can analyse seismic data with precision never seen before, leading to more efficient discovery and extraction. This is one of the most adopted use cases for AI in the energy industry, ODSU showcasing many open-source projects that are already underway in this space. The resource and mining sector have also been taking advantage of AI in a similar way.

Predictive maintenance in the oil and gas sector is another use that many have been working towards which can help predict equipment failures before they occur, ensuring operational safety and reducing costly downtime. One particular example comes from Italy’s energy infrastructure operator Snam, who are using digital technologies to make significant advancements in the country’s carbon emissions reductions.

AI in midstream operations:

  • Pipeline leak detection: Analyze sensor data and satellite imagery to identify pipeline leaks early on, mitigating environmental damage and regulatory penalties.
  • Process optimization: Optimize pipelines and refineries for efficiency and safety by employing AI to analyze operating data and recommend adjustments.
  • Smart logistics: Improve transportation and scheduling of oil and gas products by applying AI-powered route optimization and demand forecasting.

AI for downstream operations:

  • Predictive product quality: Implement AI models to predict product quality based on upstream and processing data, ensuring adherence to specifications and reducing waste.
  • Risk management: Leverage AI to assess and mitigate safety and environmental risks, ensuring regulatory compliance and worker safety.

AI for addressing price factor and market volatility

As energy pricing fluctuates significantly and regularly, AI can be used to offer some stability.

AI models can predict market trends, providing insight into global prices; information that is extremely valuable for companies and governments looking to navigate the complex geopolitical landscapes.

AI’s impact on the workforce

As AI capabilities grow, so too does the need for specific expertise within workforces to be skilled in both technology and traditional energy knowledge.

This offers a unique opportunity to oil and gas professionals to increase their knowledge base and become the torch bearers of this technological evolution.

Unions have a role in creating guidelines and frameworks to advocate for employees in these areas so the energy industry can create efficiencies with its workforce which have been very reliant on physical labour and empirical knowledge.

A real-world example is Shell’s digitalisation project that has, among other initiatives, more than 350 data scientists and over 4,000 software engineers working on developing certain AI and other digital solutions.

AI for smart grids

As the world at large demands more power, and less harmful sources of power, AI offers a chance to reduce the complexity of that task.

In the area of smart grids, the application of AI will enhance and support the two-way electricity flow using extensive data technologies. Analyzing thousands of data points, these networks produce to enable real-time adjustments. This would help the grid be agile, resilient, predictable, and improve situation awareness.

Smart grids send energy where it’s needed most; when generation is high and consumption is low, they send more electricity to storage and distribute power when usage grows and production falls. As a result, renewable energy becomes more reliable.

By using AI and insights across the industry, energy companies can find where renewable systems would produce the most energy at the lowest cost and ecological impact. This kind of informed decision-making enables a smoother, safer transition to emissions-free electricity.

For example, the US Department of Energy is awarding billions of dollars in grant funding to smart grid projects, including AI-related initiatives, to continue advancing knowledge and practical usage in this area.

AI in grid maintenance

AI presents a huge opportunity in predictive maintenance as it anticipates equipment failures by learning to identify early warning signs. These systems alert technicians to issues while they’re still easy and affordable to fix.

As a result, equipment downtime is reduced, and efficiency is increased on a level that conventional repair practices cannot match. Additionally, predictive maintenance can boost operational safety and prevent needless, and expensive, interruptions.

Utilities can minimize energy waste and disruption by keeping power networks in better condition, providing the same amount of electricity with reduced emissions.

For example, German company E.ON found that virtually inspecting their power lines with drones and AI boosted efficiency and safety. Drone-shot photos of power poles and lines are uploaded to Microsoft Azure and transferred to an AI-supported inspection tool which evaluates the images in real-time, saving time and ensuring safety standards are met.

AI for improved energy efficiency for consumers and businesses

Utilizing devices powered by the Internet of Things – a network of interrelated devices that connect and exchange data – homes, businesses, and power plants can analyze real-time conditions and adjust energy delivery in response.

This ensures as little energy as possible is used while supporting the same processes. These devices, while simple, can reduce heating and cooling usage by eight to ten percent a year, on average. Applying the same adaptive technology to larger-scale environments can yield significant energy savings.

AI plays a key role in sustainability and tracking carbon emissions but remember, AI itself consumes a lot of energy. That’s why organizations like Google, Microsoft, Amazon, and others, are pushing towards smaller models, better processing capabilities, and energy efficient hardware, while also striving towards advancements in Quantum.

For example, a new satellite, MethaneSAT, will soon be able to collect data with the help of Google’s AI and infrastructure mapping, to mitigate methane emissions. In addition to detecting emissions, the information gathered will be used to create a global map of oil and gas infrastructure with a goal of understanding which components contribute the most to emissions.

Supply chain optimization

AI has a key role to play in the reduction of larger energy supply chain challenges.

Machine learning models can analyze power networks to find areas where subtle changes could reduce emissions. For example, reconditioning power transformers would reduce the waste and emissions that manufacturing a new one would take.

In this case, AI can identify where recycling is a better path forward and make its recommendation to utilities. This alternative might be easy to overlook because of its simplicity but it can significantly impact the power grid.

Emissions reductions can also be achieved by using a closer supply, spacing shipments differently, or finding recycled material sources. AI analytics can find the best combination of these complex factors to ensure energy supply chains become as efficient as possible.

AI for weather modeling

As extreme weather events become more consequential and unpredictable, detailed forecasting and analysis is becoming increasingly useful.

Some organizations already use deep learning AI models to predict solar generation levels which vary widely in different weather conditions. This AI approach is more accurate than conventional forecasting and allows for an easier transition to effective green energy solutions.

Additionally, AI models can alert authorities of weather conditions that may disrupt power sources much earlier than conventional forecasting.

With these early warnings, power companies can ensure sufficient energy reserves and protect their infrastructure to prevent damage and outages.

For example, significant advancements are being made in weather forecasting that enable trained AI systems to make 10-day weather forecasts that are as good as traditional models and often more accurate.

AI in real-time energy trading

Another advantage of AI is that it enables faster, more profitable energy trading even as solar usage grows across the residential market and excess energy is being sold back to the grid.

If an AI system detects an upcoming surplus of solar energy, it can automatically trigger mechanisms to sell excess energy back to the grid or divert it to storage systems for later use, enhancing the overall efficiency of energy trading operations.

For example, GE’s Alpha Trader, a market forecasting software, enables real-time and day-ahead analytics using digital twins as well as AI and machine learning.

Enterprise AI at scale

Shell have been using machine learning across assets in exploration upstream, midstream, downstream, and retail to drive efficiencies across the business leading to billions of dollars in savings.

They’ve also created an Open Energy Initiative with C3 AI, Baker Hughes, and Microsoft, a first-of-its-kind open ecosystem of AI-based solutions for energy and associated process industries.

This provides a framework for energy operators, services, and equipment providers to offer interoperable solutions, including AI- and physics-based models as well as monitoring, diagnostic, prescriptive actions and services for energy use cases.

Chevron, BP, and Exxon have their own programs for AI-enabled control towers, using technology like digital twins to accelerate towards their “industrial metaverse” vision (which is a topic we’ll be covering in the near future).

Another example is Palantir, who are accelerating AI transformation in the energy industry through customized data integration, modelling, and simulations to deliver interactive and real-time information to help with decision-making.

How to be successful in using AI

Having a solid foundation for AI implementation is key to long-term success. But what does that really mean?

First, it’s important to address data, data governance, and implementation challenges in the energy industry to identify the best approach.

Data quality and availability: The industry depends on reliable and accurate data for various operations such as safety, compliance, system reliability, and customer service. Data quality can be compromised by noncompliant and mismatched information from different sources, weak controls, and data silos.
To overcome this challenge, energy companies need to implement strong data governance strategies to ensure data quality, consistency, and accessibility across the organization.

Data security and privacy: The energy industry is vulnerable to data breaches and cyberattacks that can jeopardize the security and privacy of data. Data governance can help protect from unauthorized access, use, and disclosure by establishing firm policies, standards, and procedures. Consider using data clean rooms and data officers to balance the need for data sharing and collaboration with the respect for privacy and trust.

Data analytics: Effective use of data analytics can optimize performance, reduce costs, enhance the customer experience, and drive innovation. This involves high-quality and relevant data as well as skilled and experienced data professionals. Data governance can support data analytics and AI by providing data cataloging, lineage, and metadata management, as well as data ethics and stewardship. It can also help align data analytics and AI initiatives with the goals and values of the organization.

How to get started

It’s a significant undertaking to commit to leveraging AI within an energy company but the long-term benefits will yield significant positive results if done correctly.

To get started you’ll first need to answer these questions:

Define your AI vision and objectives: What are the problems you want to solve with AI? What are the expected outcomes and benefits? How will you measure the success of AI?

Assess your data readiness and maturity: What are the sources and types of data you have? What is the quality and availability of the data? What existing data governance practices and capabilities do you have in place? How can you improve your data readiness and maturity?

Identify and prioritize your AI use cases: What are the specific applications of AI that can address your problems and objectives? What are the data requirements and feasibility of each use case? What are the risks and challenges involved? How will you prioritize and select the most promising use cases?

Develop and deploy your AI solutions: How will you design, build, test, and deploy your AI solutions? What are the tools and technologies you will use? What are the skills and resources you will need? How will you integrate your AI solutions with your existing systems and processes?

Monitor and evaluate your AI performance: How will you track and measure your AI solutions’ performance? How will you ensure the reliability, accuracy, and fairness of your AI solutions? How will you handle the feedback and issues from AI users and stakeholders? How will you update and improve the AI solutions over time?

A solid data foundation for embracing AI

AI is a key player in the future of global energy. Its potential to revolutionize utilities’ operations, and its pivotal role in promoting renewable energy, cannot be understated.

As we navigate the challenges and opportunities of this era, the energy sector stands on the brink of a new age of efficiency, safety, and sustainability.

The key to successful AI implementation is undeniably a solid data foundation and successful bridging of the information technology and operational technologies while embracing AI investments.

Consider all the data sources that are intertwined in the energy sector: SCADA, AMI/Smart, GPS, drone imagery, LiDar, telemetry, ERP data, external data such as weather, etc. Now, you have to collect, classify, organize, monitor, secure, govern, analyze, predict, simulate, and interact with this data to get the outcomes desired.

While the tools designed to simplify this complexity and technology exist today, technology providers will keep investing and introduce more of them to increase their cloud revenues. While generative AI and LLMs from Microsoft, OpenAI, Google, and more simplify these complexities and accelerate AI adoption, the data foundation, ontology, and the narrow AI needed for your energy business can’t be supported by these alone.

Generative AI, which is at the center of AI usage today, can certainly help in the consumer and service space, providing support to desk-based workers, HR, improving client interactions, training staff, and disseminating large amounts of data, generating content, or boosting productivity of frontline operators. But most high-value AI will still be predictive and doing simulation for “what if” scenarios in the energy industry. With your business well integrated to interact with Generative AI assistant, you can get to the next level of AI transformation for your business and its processes.

Reaching the level needed to deploy predictive machine learning requires high-quality, well-labeled data for each use case and context. Building predictive AI models with generative AI models can help but does not entirely solve data quality challenges.

Connect with us to get started

Our team of dedicated professionals can help you determine which options are best for you and how adopting these kinds of solutions could transform the way your organization works. For more information, and for extra support along the way, contact our team.