The energy sector is a highly technology-driven industry. With needs to handle natural resource data among large pieces of equipment in harsh conditions, the oil and gas industry has long used data methods and various technologies to make processes more efficient. More recently companies in the energy industry have begun to ramp up their adoption of various AI technologies to help in a variety of ways including ways to make our energy consumption more efficient. With the advent of widespread access to big data technologies, low-cost compute resources, and the increasing availability of technology that implement the seven patterns of AI, it’s making it easier for the energy sector to see real value from AI and ML.
In heavily regulated industries such as the energy industry there are a number of unique challenges to AI adoption. In a recent AI Today podcast Dr. Satyam Priyadarshy, Technology Fellow and Chief Data Scientist at Halliburton shared his insights on how the use of data has changed in the energy industry over the past decade, some use cases for how AI and ML is currently being applied, as well as how county level strategies are having an overall impact on AI. In this follow up interview he shares his insights in more detail.
How is AI currently being applied in the energy industry?
Dr. Satyam Priyadarshy: The energy industry has been implementing data science and AI solutions in all aspects of the business lifecycle, with varying degrees of success in the past. However, with the advent of easy access to big data technologies, their scalable implementations and deployments are increasing in the energy industry. For example, using the video analysis obtained from drones, in real-time, to look at leak detection of pipe, amount of dirt accumulation on the solar panels, or the bend in large blades of windmills. We have pioneered the development and deployment of AI solutions based on modified natural language programming algorithms on unstructured data of the oil and gas industry to reduce capital waste and build actionable insights in near real-time. Over 100 business cases have been targeted for the energy industry that leverages simple clustering to complex deep learning algorithms, with varying degrees of economic value generation. One of the key factors in our success has been the development and deployment of cloud platforms like iEnergy (the oil and gas industry’s first hybrid cloud solution) and development platform with open access for the industry- OpenEarth.community.
What are some challenges around AI adoption in the energy industry?
Dr. Satyam Priyadarshy: FEAR describes the challenges that the energy industry faces when it comes to the adoption of Artificial Intelligence and Data Science at large. Here, FEAR stands for the following four key challenges:
First-principles, science, and engineering have dominated the industry for a very long time, and it’s been hard for many professionals to think and implement data science and AI solutions at scale.
The evolutionary pace of emerging technology is incompatible with the adoption of those technologies in the industry. The tell-tale sign of the gap in adopting search technology is an example of why significant time and resources is wasted in searching for the data sets needed to build the data science models, thereby reducing the impact of AI in the industry.
Accomplishments of the past shadows the adoption of emerging solutions, very evident from comments one hears in the industry, that we have been the pioneers of high-performance computing and large volumes of data.
The reactive nature of addressing breakdowns, non-productive time, and other operational aspects have been long-standing practice and culture. Data Science and AI enable the proactive culture to address the possibilities of inefficiencies creep ups and thus requires a transformational change in the industry.
How have you seen the use of data change in the energy industry over the past decade?
Dr. Satyam Priyadarshy: The energy industry has been a creator of large amounts of multi-dimensional, multi-variate, and diverse data sets for decades. However, the maximization of value from the data remains a challenge even today. In the last decade, the maturity and easy access to Big Data technologies, Cloud computing paradigm, and platform approaches have made great strides and progress towards leveraging the data that the industry has. However, it is nowhere close to the ‘data-native companies’ in terms of optimizing and maximizing the value.
What are some of the challenges for dealing with data and AI from an oil and energy industry perspective?
Dr. Satyam Priyadarshy: In March 2015, CNBC carried a story titled Oil firms are swimming in data they don’t use based on study by the McKinsey & Company. The key message in the story and study was that the oil and gas industry used only 1% of the data they collected, where the desire by the executives was to leverage 95% of the data. A huge gap between what is in practice and what is needed? Unlike other industries, the data related challenges are complicated and complex for the oil and gas industry. If we look at the industry level, the data democratization and sharing of knowledge from data-driven innovation have been very limited to the point of no sharing. However, if we look at a company or enterprise, the existence of data-silos and cultural silos has prevented us from leveraging scalable AI-Driven innovative solutions to generate value from the data.
How do large organizations approach change management for technologies such as AI?
Dr. Satyam Priyadarshy: Data science and AI have proven enormous strategic and economic potential across all industries. Hence any organization large or small can take advantage of the maturity of AI applications, however, it requires a comprehensive integration of Data Science and AI into products, services, workflows, and business models. For integration success and growth of the business, three transformation areas of the organization become critical and they are: (1) A comprehensive knowledge of data science and AI in the right context, (2) A strategic change at the top leadership level, (3) A framework for success to achieve automation, optimization, and innovation, (4) A talent readiness approach and (5) finally, a technology-agnostic view for implementing and scaling data science and AI solutions.
How is the energy industry approaching data related issues around security, privacy, transparency, and ethics?
Dr. Satyam Priyadarshy: The energy industry is traditionally a highly regulated and compliance-focused industry. Hence the industry has very mature data governance and security policies in place, to an extent sometimes it poses a challenge to use the data for internal model development and research purposes. As the maturity increases about the knowledge and impact of data science and AI, the industry players are modifying and revisiting some of the more restrictive aspects of data governance, while deploying solutions for real-time monitoring of data access, transparency, and ethics. What are the best practices in the era of AI deployment, is a topic of interest and discussion across all industries, and with every passing month, more and more information is available and practices adapted in an agile fashion to maximum value from data, while continuing to decrease the risk associated with the use of data, while increasing the safety, privacy and ethical aspects of data-driven innovations?
How is the role of data science and data science teams changing at large organizations?
Dr. Satyam Priyadarshy: Data science is best understood simply, to do science on the data. To do experiments in science, one leverages a multitude of tools, technologies, hypotheses to ask questions and find answers. Similarly, data science leverages the first principle scientific and engineering solutions, simple data mining and statistical methods, and data-driven innovation using scalable AI solutions. For over 6 years I have been leading the oil and gas industry’s first center of excellence for big data and data science, which generated significant value for internal and has generated interest, value, and implementation of data science and AI solutions at multiple organizations located in many countries. What we have pioneered and proven is that co-innovation approach among the data science teams, the domain or subject matter experts, and the business leaders, to achieve significant economic value for the organizations. We have developed a SMART DigitalRM approach to educate, engage, and empower data science and business teams to achieve success with Data Science and AI solutions.
What do you see as critical needs for workforce development around AI?
Dr. Satyam Priyadarshy: There exists 100’s of courses on AI-related topics, and they are great for mass awareness of the topic. However, the practice of AI and data-driven innovation requires talent transformation, in the right context with the right content. From our six years of engaging the talent transformation across the globe for the oil and gas industry, where we have conducted highly contextualized talent transformation workshops, boot camps, and masterclasses for the industry professionals, and have been measurable and quantifiable value creation by the trained talented compared to generalized workforce development courses, etc. Another area that we have to focus on is Executive training for the new era. Since hiring, managing, and retaining data scientists is culturally and organizationally different than the traditional oil and gas or energy industry practices.
You’ve also been involved in helping the country of Mauritius develop their AI strategy. Can you provide more details around this?
Dr. Satyam Priyadarshy: Yes, in 2019, I was humbled to be invited to be a member of the Mauritius Artificial Intelligence Council, the Republic of Mauritius. As one of the advisors, my role is to provide insights into how to achieve success in leveraging data-driven innovation and AI for the various initiatives that are part of the vision of Mauritius for the next decade. The council’s role is to help formulate an action plan, propose and evaluate the right AI solutions for the benefit of Mauritius, its national economy, socio-economic activities, and the environment for top AI-talents and companies besides others. The council constitutes a large number of top leaders within Mauritius and others located globally.
What AI technologies are you most looking forward to in the coming years?
Dr. Satyam Priyadarshy: Artificial Intelligence should be viewed as an innovation enabling field more than just a technology. The application of AI will impact all aspects of work and personal life in the years ahead. In the coming years, applications of AI in three focus areas will be important in transforming the businesses in the era of compounded disruption. The three focus areas are automation, optimization, and innovation. For example, to have a digital twin for a drilling workflow requires an integration, assimilation, execution, and simulation of digital twins for components on the platform, to be automated. The workflows in the energy industry are complex and present multivariate challenges, so optimization using AI becomes very valuable for increasing efficiency and productivity. As emerging technologies evolve, their deployment for different workflows will require the development and deployment of field-ready innovative solutions that are data-driven AI-enabled.
In summary, data science and AI will provide numerous solutions to make businesses resilient, sustainable, and safer enabling them to navigate the compounded disruption landscape. However, this requires organizations to overcome FEAR and take advantage of SMARTDigitalTM approaches.