• In our previous post, we discussed how important situational awareness is for drilling automation. Full situational awareness requires the use of not one, but many models – both physics based as well as data-based (AI). Are there strategies that one can employ to get the rig crew to trust AI?

    The rig crew in general would not change their operational workflow unless they can be shown clear and consistent benefits using a new tool. A known problem with AI models is that even when fully validated through use in over 1000s of wells, they may produce a bad output when faced with a unique scenario not seen before. Embedding physics in the AI model makes it less likely to falter, but model retraining may still be required.

    It certainly also helps if the AI model is interpretable. That will help explain the output of the AI model to the rig crew if the outputs are not as expected. If there is an issue of bad output, the rig crew would want a solution ASAP and will be less tolerant of long delays to a solution. An agile software update workflow is not useful, but essential.

    The software tool provider should consider each rig crew involved in operations to be an important stakeholder who can influence the adoption of the AI based drilling advisory. Often, stakeholder management on a rig is done in an ad-hoc manner. However, if this process is diligently approached, it is one of the best paths to change management. This involves considering each rig crew involved in the drilling operation and identifying their level of interest and influence (Slide 3). The tool provider needs to respect the fact that where a person is on this chart is very much dependent of that person’s prior experiences (or lack of) using such tools.

    The end goal is shown in Slide 4. It is to help move all the users to right side of the chart. As one can imagine, enabling trust in AI is not easy and will take time. 

    Click below for the slides on this topic:

    Change Management