Autonomous Decision-Making While Drilling

The drilling process is complex because unexpected situations may occur at any time. Furthermore, the system extremely long and slender, therefore prone to vibrations often being dominated by transient periods. Adding fact that measurements are not well distributed along system, with majority of real-time only available top side having access very sparse data from downhole, poorly observed making it difficult use standard control methods. Therefore, achieve completely autonomous operations, necessary utilize a method capable estimating internal state parsimonious information while able make decisions will keep operation safe but effective. A solution enabling decision-making has been developed. It relies on an optimization time reach section total depth (TD). estimated TD decomposed into effective spent in conducting likely lost solve events. This problem solved using Markov decision method. Several example scenarios have run virtual rig environment test validity concept. found adapt itself various conditions, as for aggressive when runs smoothly uncertainty states low, also more cautious downhole conditions deteriorate or observations tend indicate erratic behavior, which prior event.