_ When it comes to drilling automation, deciding where your technology should be deployed is a critical factor in overall solution design. Similar to other industries, there is an eternal debate on the merits of cloud vs. edge systems. Cloud systems benefit from having scalable computing, easy access to historical databases, and potentially global teams of experts providing support; however, they require maintaining a high-speed, high-uptime networking solution to the drillsite to be effective. Edge devices, on the other hand, can immediately react to real-time data that is being produced by the rig regardless of network status, though they are limited by the personnel and hardware resources that are in place on location. For drilling in particular, there is yet another nuance to consider: While edge devices are closer to the action than cloud, they may still be several miles and an additional low-bandwidth communication link away from the where the magic really happens, the drill bit. This leaves us with yet another place for a drilling automation solution: downhole. Here trade-offs become extreme—hardware in small form factors that must be ruggedized for the drilling environment and autonomous operations outside of a limited downlink command set. There is no single right answer to this problem. With this month’s paper selections, we will explore possibilities for matching the right location with the right automation strategy. In paper SPE 224743, the critical challenge lies in precise geosteering for maximum reservoir contact. The decisions being made are at the larger scale, replanning the well’s trajectory in real time. Successful operation requires connections between cross-functional teams and systems in a task best performed in a remote operations center supported by a cloud system. The models of geology and the drillstring are used to distill a minimal command set that can be sent back to the rig and ultimately transmitted downhole to a rotary steerable system. Moving on to paper SPE 227896, the scope of the task has narrowed. The near-term well plan is fixed, but the drilling parameters must be managed when drilling a standdown using a steerable motor assembly. In this case, a physics model of the drillstring is loaded onto an edge device, enabling the rig control system to infer what is happening at the bit and automatically deliver efficient slides with minimal manual intervention from a remote center. Finally, in our last paper, SPE 225656, the focus moves to immediately behind the bit. Information about the well plan and drilling parameters that will be used is loaded directly into downhole tools, which then execute the steering. If the drilling process proceeds within that envelope of constraints, the rotary steerable can independently respond to downhole disturbances minute-to-minute in a fashion that would not be practical from the rig floor.
Marc Willerth (Sun,) studied this question.