Artificial intelligence (AI) is moving into practical industrial applications, from predictive drilling and production optimisation to automated manufacturing and supply chain management.
ADIPEC speakers stressed that scaling adoption requires more than new software, calling for redesigned processes and investment in data infrastructure, ownership and workforce skills.
Girish K. Saligram, President and CEO of Weatherford, described AI as “the most profound change that all of us are experiencing in our lifetimes. It’s not just a technological change; it’s a societal revolution,” highlighting its growing impact across the oil and gas value chain.
Applications in drilling are helping operators anticipate subsurface conditions, identify potential problems, and improve decisions on where to drill. AI is also enabling the reinterpretation of seismic data and more effective production optimisation. Saligram said AI is already improving productivity in Weatherford’s service operations, with tasks once requiring four to eight people now performed autonomously, although industrial AI remains at an early stage.
AI is the most profound change that all of us are experiencing in our lifetimes. It’s not just a technological change; it’s a societal revolution.
For Siemens President and CEO Dr Roland Busch, the challenge is moving beyond isolated pilots to embed AI across industrial processes. He argued that simply adding a digital assistant to existing operations would deliver limited benefits. Instead, companies need to fundamentally rethink how their processes work and integrate AI into their underlying systems. “If you put a copilot on top, it stays shallow. You really have to rethink your processes deeply, change them, and then throw AI on top of it,” Busch said.
He highlighted the importance of training models on specialised industrial data, which differs considerably from the language-based information used to train conventional AI models. Access to engineering, design and operational data will be essential, although concerns over intellectual property (IP) continue to make companies reluctant to share information.
The commercial case, however, is becoming increasingly compelling Busch predicts that AI could boost plant productivity by 20%, halve product development cycles, and improve energy efficiency and manufacturing flexibility.
Busch then linked industrial AI to supply chain resilience, arguing that automation and robotics could make local manufacturing more competitive by reducing dependence on labour-intensive production in overseas markets.
The question of who controls the data underpinning these technologies was central to Uljan Sharka, Founder and CEO of Domyn. He argued that AI sovereignty is fundamentally about data ownership and the ability of organisations to control their technological future.
“AI is not software, AI is data,” Sharka said, warning that once company information is incorporated into an AI model, it cannot simply be extracted. He compared AI to electricity, arguing that organisations which depend entirely on external providers risk losing control over critical operations.
Sharka advocated a more decentralised AI model, combining open models with independently controlled infrastructure. “We need to change the way we build AI. We need to make it more decentralised, and we need to put companies in control of their own data.” He pointed to a four-billion-parameter language model trained on a European supercomputer in just 45 days as evidence of growing accessibility.
However, he cautioned that ownership of an AI model alone does not guarantee sovereignty if companies remain dependent on a handful of infrastructure providers.
Accountability and workforce development will also shape adoption. Sharka argued that AI should not be treated as an autonomoustead, AI models must be combined with deterministic data science and certified information sources to ensure greater control and reliability
Busch acknowledged that automation would eliminate some repetitive and rule-based roles, but argued that it will also create new opportunities, including robotics programming and automated operations. This will require substantial investment in retraining. “We have to rebuild our workforce, and that means we have to retrain it and train it on new technologies,” he said.
The panel agreed that industrial AI’s value will depend on practical deployment, with Saligram stressing the need for reliable data and accountability. Realising its potential will require aligning technology, people, and infrastructure around measurable operational gains.
KEEPING THE ENERGY INDUSTRY CONNECTED
Latest Articles
ADIPEC 2026 Day 2 puts infrastructure delivery at the centre of efforts to meet future energy demand
Korea Pavilion brings manufacturing excellence to ADIPEC 2026
Capturing Egypt’s huge energy hub promise involves earning long-term capital confidence
What it takes to make cross-border collaboration viable and scalable in a new reality
Watch/Listen
Building the future of green molecules with platform-based engineering
In an exclusive Energy Connects studio interview at ADIPEC 2026, Hengwei Liu, CTO and Head of Green Hydrogen of Wison Engineering, explains why a new…
Schneider Electric: turning AI and autonomy into real-world energy performance
In an exclusive Energy Connects studio interview at ADIPEC 2026, Georges El Mir, Senior Vice President, Industry & Process Automation, Middle East & Africa at…
How gas, renewables, and innovation are shaping the Middle East’s energy future
Ahead of ADIPEC 2026, Monaam Htiouech, Vice President and Head of Gas Service Sales, Middle East at Siemens Energy, discusses how rising demand from population…
Why energy and technology must work together to fuel future growth
Ahead of ADIPEC 2026, host Danielle Martin speaks to Amos Hochstein, Managing Partner at TWG Global, in the latest episode of the Energy Connects Podcast…
