Autonomous drilling demonstrated – Breakthrough research and development on autonomous systems and AI at NORCE

We are very grateful for the great support we received from our partners in the project. The successful demonstration of autonomous drilling is a direct result of the good collaboration and teamwork in the project.

By autonomous system here we mean the highest level of automation, namely a system that is able to adapt to changing environments and changing goals, to learn both from experience and on-the-fly, and to make appropriate choices given perceptual limitations and finite computation.

The autonomous drilling system has embedded protection functions such as safe operating envelopes and fault detection, isolation, and recovery functionalities, that ensure its safe use. With the latter-mentioned safety triggers, the system is capable of detecting, automatically abnormal situations and to apply mitigation procedures. Within the project, we have extended these safety triggers with automatic functions that attempt to cure the original cause of the problem in order to put back the operation in a normal drilling state.

At the same time, within the project we address potential situations that might require to release control from autonomous to manual mode and thus we implemented into the system a management of safe mode transitions.

In this way, we address one important question related to human factors in automated/autonomous system: how to ensure a safe transition from system-controlled mode to manually controlled mode when the situation awareness of the human operator may be very low?

Our solution always takes the current context into account. The autonomous system estimates suitable states which enable the drilling operation to be safe for a few ten seconds and performs transitions to these states if necessary, in order to leave sufficient time for the human operator to take control and handle the current situation.

In drilling, there is only one way to reach the end of a well and that is to drill stand after stand. But the time spent to drill each stand can vary and there may be intermediate actions to perform that are not directly contributing to increase the borehole depth, some are planned operations and others must be executed to either prevent a problem from occurring or cure an issue which has already happened.

The decisions that need to be taken in the context of autonomous drilling are not only about optimizing the speed of drilling but also to perform actions that might not look optimal in the short-term perspective and yet, they would decrease the risk of occurrence of drilling events in the long term.

More about this: Autonomous Decision-Making While Drilling (mdpi.com)

The research and development of an autonomous drilling system is conducted as part of a NORCE-led project, financed by Norwegian Research Council and AkerBP, Equinor, Petrobras, Repsol, TotalEnergies and Vår Energi, through Demo2000 program. This is a joint work between NORCE, Sekal, NOV and Odfjell Drilling. The project is entitled “Demonstration of Autonomous Drilling Process Control” (DADPC) and the demonstration was performed on NORCE’s full-scale rig, Ullrigg.