Oil and gas companies are becoming increasingly eager to scale up AI implementation across their operations, drawn by the potential to speed up technological development, improve efficiency and ultimately improve the bottom line. But moving beyond individual use cases to wider deployment requires careful planning, as a new briefing note from consultancy EY points out.
EY highlights the benefits of scaling up AI platforms, despite the high cost, and identifies the challenges in ensuring it is done effectively and that these platforms deliver consistent benefits.
“Evolving from proof of concept on small projects to scale is a complex undertaking, and it’s the last mile that proves most daunting,” EY says in the briefing.
The consultancy identifies key areas on which companies should focus if they are to scale up AI implementation successfully, putting human capabilities to the fore. These echo themes identified in lively discussions at recent Future Digital Twin & Generative AI conferences in Abu Dhabi and Amsterdam, as well as other research we have reported on recently.
Top of the list is that buy-in from and active involvement by the company leadership is vital. EY says that as AI becomes more ubiquitous, it will require access to data from multiple systems across a company, potentially creating turf issues and integration challenges. It says the corporate leadership has a key role to play in making sure that different functions work well together and in supporting the development team.
The higher stakes involved in signing off on AI investments to enhance subsurface operations where millions of dollars are on the line, compared to back-office operations are also highlighted.
Abhilash Krishna, Manager, Technology Consulting, at Ernst & Young LLP says that companies need to clearly analyse the risk of various AI initiatives and weigh them against measurable benefits.
“While reservoir simulations and drilling functions may face higher risks to using AI, they also have the opportunity for more significant rewards. Leaders who embrace that risk — and give their people permission to fail — will solve the AI puzzle sooner and gain an advantage,” he says.
The second key topic EY says companies need to address is the increased operational complexity that accompanies the scale of up of AI. More robust and insightful AI requires more robust, insightful information to feed it. That means personnel requirements, processes and technology also need to scale exponentially.
One aspect of this scale up that will require careful consideration is whether to use third party vendors or develop in-house functions to handle leading opportunities to create value from AI and emerging technology, such as predictive maintenance for heavy equipment and assets and intelligent optimization of operations performance.
“Making the decision around what to build versus what to buy will be more important than ever when it comes to the innovation around AI technology. Ultimately, you may find that relying on outside resources to supplement internal staff can speed up your AI efforts, because you’ll have more total resources to apply to development and implementation,” according to EY.
Cultural shift
The final issue identified by the consultancy is the need to overcome cultural resistance, which it says could be the biggest risk that oil and gas companies face in covering the “last mile” in implementing AI.
Just building AI tools is not enough — people must adopt and maintain them. Change management focusing on communicating AI strategy and benefits and how work will be transformed is vital, EY says.
“You can’t keep AI strategy and development activities under wraps and then expect employees to embrace these tools when they come online,” said Himanshu Dabral, Senior Manager, Business Consulting at Ernst & Young LLP. “It’s important for companies to view their AI efforts as something greater than an IT project, because at some point they will need employees across the enterprise to adopt and adapt.”
Following on from this, employee input and feedback is crucial in maximising the success of AI implementation. “Did AI give correct answers? Did it save time or reduce busy work? Is there data that is missing or incomplete? AI calibrates itself and gets better over time, but it requires feedback — and constant data maintenance — to do so,” EY says.
(Photo: Shutterstock)
