eDrilling is introducing what it says is a fully adaptive, self-learning and continuously improving system that addresses the challenges and requirements of drilling. Called AI Drilling Agent, it finds the most efficient and safe way to TD, predicts incidents and hazards, and optimizes parameters.
“The AI Drilling Agent project is the biggest science and engineering effort we’ve undertaken as a company,” says Toni Fadnes, eDrilling CEO. “We have always demonstrated leadership in AI R&D and developing an everyday companion to engineers and crew members will make AI real and relevant for our industry globally.”
eDrilling is calling the agent “a copilot, a wingman, a companion to your drillers, drilling engineers and crews, it be geothermal, CCUS or legacy hydrocarbon operations.” The technology builds on previous research and engineering breakthroughs like dynamic modeling, digital twins, model-based reasoning, autocalibration, and dynamic simulation.
The AI Drilling Agent has status awareness as it detects operational status, including rig activity detection and rig state tracking. eDrilling stresses that it is a learning machine, not a database. The Agent does not retain access to data used or analyzed in operations or training – just the knowledge. “Like a teacher, who through extensive study is enabled to advise and explain, without having all the data memorized,” according to eDrilling.
Further, the technology is technology, vendor and operation agnostic so it can be deployed on any rig and any rig.