Today's Edition
BATOOL GHAITH (ABU DHABI)
Artificial intelligence is becoming increasingly embedded in energy operations, helping companies predict equipment failures, optimise refinery and production performance, monitor energy use and emissions, and reduce the need for workers to enter hazardous environments, industry experts said.
According to experts, across oil, gas and industrial facilities, the strongest results are emerging in areas such as predictive maintenance, process optimisation and continuous monitoring, where AI can use large volumes of operational data to identify inefficiencies or developing problems before they interrupt production.
Liam Hurley, President of Middle East and Africa at Emerson, said energy operators are moving from reacting to equipment and process problems to anticipating them.
“Industrial AI is moving beyond monitoring performance to actively preventing inefficiency, detecting energy losses, equipment deterioration or process flaws well before they become visible,” Hurley told Aletihad.
He said predictive maintenance is one of the areas where operators are seeing returns first, noting that equipment that was previously difficult to monitor continuously, including steam traps, relief valves and remote wellheads, can now generate data that analytics platforms use to identify potential problems and prioritise them according to risk.
This allows operators to intervene before equipment deteriorates to the point of failure, potentially reducing unplanned shutdowns and unnecessary maintenance, Hurley added.
He also indicated that some of the clearest measurable gains are coming from process optimisation and planning.
At one of the region’s leading energy companies, he said a hybrid model combining first-principles engineering with industrial AI achieved yield and quality prediction accuracy of up to 98.5% in key refinery units.
That allowed planners to make feedstock and blending decisions with greater confidence and reduce the need for costly manual adjustments, Hurley noted.
He added that there are measurable returns in energy management.
“Energy is one of the largest controllable costs in any facility, and analytics that track consumption continuously can pinpoint losses as they happen, so savings show up directly in operating costs and emissions intensity,” he added.
“The operators seeing the biggest returns treat AI as part of everyday operations rather than a pilot, and scale proven use cases across units and sites,” Hurley said.
Predicting Problems Before Production Stops
Dr Hamdi Sheibani, Dean of the College of Engineering at Abu Dhabi University, also identified predictive maintenance and production optimisation as among the strongest current applications of AI in energy.
He said these areas can improve equipment reliability, support more consistent output and improve energy efficiency. Leak detection and emissions monitoring are also becoming increasingly relevant, allowing operators to identify and address abnormalities earlier.
“The greatest gains come when these applications work together, giving operators a clearer picture of performance and helping them deliver energy more efficiently, safely and sustainably,” Sheibani told Aletihad.
Over the next five years, he said AI is expected to have its greatest impact on the daily operation of producing assets and processing plants.
Connecting information on production planning, equipment health and energy consumption could allow engineers to optimise entire facilities rather than individual pieces of equipment, he added.
“With stronger data foundations and workforce skills, AI can support more reliable energy supplies, higher productivity, and lower emissions per unit of output,” Sheibani noted.
Robots Enter Energy Inspection
AI is also changing how energy facilities can be inspected. Dr Abdalla Swikir, Assistant Professor of Robotics at Mohamed bin Zayed University of Artificial Intelligence, is developing autonomous robotic systems that can conduct repeated inspections, collect visual and thermal measurements and identify changes in equipment condition.
“The aim is for robots to handle this repeatable work so that engineers can spend more time investigating problems and deciding what action to take,” Swikir told Aletihad.
For example, an increase in equipment temperature may be expected when the workload increases, but the same increase under otherwise similar conditions could warrant investigation, he noted.
“The role we see for AI is to help interpret these observations in context, rather than simply generate an alert whenever a reading changes,” he said.
Visual and thermal sensing can provide different information about equipment conditions, while autonomous systems could eventually help monitor areas that are difficult or hazardous for workers to access, according to Swikir.
Industrial validation would also need to address hazardous area certification, integration with maintenance procedures and what happens when a robot cannot confidently complete an inspection, Swikir explained.
Safety is another major area where AI and automation can change energy operations. Sheibani said AI-assisted robots and drones can be used to inspect offshore structures, confined spaces and high-temperature equipment.
Hurley similarly said continuous monitoring is replacing some manual inspections, while robotic handling and AI-powered machine vision are increasingly being used in industrial settings for repetitive or higher-risk tasks.
On emissions, Hurley also said the same operational data used to monitor equipment performance can give operators a more continuous view of energy use and emissions rather than relying only on periodic estimates.
