Automated Planning of Procedures for Risk Mitigation and Efficiency

Abstract

Abstract To effectively manage drilling operations is an enormous challenge. The work must be carefully planned and scheduled, but unforeseen events may force changes in the plan at any time. An efficient plan of operation is vital to ensure that work is performed safely and efficiently, that risks are managed appropriately, and that the cost of the operation is controlled. Maintenance tasks must be performed on schedule; if necessary maintenance tasks are delayed the costs associated can balloon or health and environmental risks might arise. However, routine maintenance must be scheduled to minimize the impact on ongoing operations. The National Aeronautics and Space Administration (NASA) is currently pioneering work on using automated planning and scheduling to manage the work performed during Extra-Vehicular Activities (EVAs) on the outside of the International Space Station (ISS). These EVAs are the most difficult activities performed by the astronauts on the ISS, and each one is planned to maximize efficiency and minimize risk. The time available for an EVA is strictly limited both by the life-support system and human endurance, and every minute must be utilized properly. NASA is developing a technology for representing procedures, or the tasks that must be performed during the EVA. The procedures encode the time estimated for the task, the tools and equipment needed, any preconditions for the task to be performed, and also the state of the system at the end of the task. An automated planner has been developed that can take a high-level goal plus those procedures, and: produce a plan of operation that attains the goal while ensuring the tools and equipment are available when needed, ensure limited resources such as oxygen are managed efficiently, ensure the safety equipment is not overloaded, and ensure that no performance rules are violated. The system can track the usage of tools and equipment and automatically update inventory databases at the end of the EVA to represent changes in location, to simplify the beginning of the next EVA. In the event of unforeseen events, the automated planner can re-plan the remainder of the shift. This planning technology developed for NASA has potential to improve the safety and efficiency of drilling operations. Risks associated with safety-critical activities can be reduced by validating the sequencing of these activities to ensure there are no missing steps or misconfigured systems that could lead to accidents. Compliance with safety policies can be improved by modeling these policies as constraints on how drilling activities should be performed. Logistics for drilling preparation and operations can be improved by automatically deriving the tools and equipment needed based on the tasks to be performed. This paper will describe the planning approach implemented for NASA, and will identify how that approach can be used to improve drilling operations.