Bridging the operations gap: Bringing AI to the edge

As the energy transition accelerates, the shift away from new oil and gas exploration intensifies. For operators, that reality puts a premium on maximising production and recovery from existing assets. AI solutions are playing a key role in helping oil and gas operators realise value and minimise CAPEX inefficiencies on the planning side. That’s been less the case in operations, where field environments can be harsh and disparate, data is often siloed, and communications between the office and the field can be a challenge. Bringing AI to operations was not possible. Until now.

The imperatives to produce more with less, minimise operational costs, improve safety and reduce emissions are simply too compelling an opportunity. We’re now at an inflection point in bringing AI into operations, bridging the divide between office and field and between planning and operations. That’s where the next tier of returns from digital will come from in our industry.

Think about today’s conventional operations. Reservoir, petroleum and production engineers generate a lot of insights in the office desktop environment. Far too often these insights fail to get executed in the field. In a not an uncommon scenario, an engineer in the office makes a call to an operator in the field to change a setpoint. Because of inertia, pushback or miscommunication, it may never happen.

This disconnect between planning and operations can result in unplanned downtime due to equipment failure or nonproductive time spent troubleshooting. It makes it harder to stay ahead of production challenges, whether it’s paraffin deposition or high-water cuts in wells.

The solution: distributed AI in both the planning and operations environment, conditioned by physics-based simulations that provide trust in the AI predictions and bridging these environments through technology integration between the edge and the desktop.

The most illustrative analogy is probably the autonomous vehicle. It requires two infrastructures working together. At the edge, the compute infrastructure, sensing and AI model reference data in real time to drive the car in the correct lane on the right navigational path. The second infrastructure is looking at all the other cars on the road along the route, measuring their speed, assessing traffic and road conditions, crunching all those billions of data points in the cloud and sending that intelligence back to the car itself to select the best route. It’s this partnership between the edge and the cloud that makes this autonomous and optimal journey from point A to point B possible.

We’re seeing something similar playing out in oil and gas production operations. In the desktop environment, whether it’s public cloud, sovereign cloud or on-site, science- and physics-based simulators can model reality and bring the all-important trust factor to the underlying data. Fusing those simulations with AI capability to identify trends from that data creates much better and faster insights in real time.
Those insights are then sent back to the edge environment, near the wellbore, the rig or the pipeline, to facilitate autonomous operations. The result is better insights, trust in those insights and seamless execution in the field — without the risk of instructions going awry.

SLB is the only company in the industry doing this right now. Working closely with an operator in South America, we deployed a smart production operations solution in a remote brownfield that increased oil production by 4%, dramatically reduced production losses, raised crew efficiency by over 60%, reduced well failures by 25%, lowered CO2 emissions by 57%, and boosted chemical treatment reliability to 99%.
We’re also working with a major operator in the Middle East on a similar concept, deploying our OptiFlow™ production assurance solutions and Agora™ edge AI and IoT solutions. These solutions combine real-time intelligence at the edge with full-field visibility across wells and gathering networks. This will enable autonomous capabilities that reduce uncertainty between planning and operations and significantly increase proactive actions that improve production performance.

Ultimately, what these technologies enable us to do is bring the entire production community into a single operations environment. Wells, pipelines, facilities, operations, maintenance and all the other elements of the fragmented production landscape are finally able to come together as an integrated and collaborative one. That’s a game changer for the industry.