A New Approach to Development and Validation of Artificial Intelligence Systems for Drilling

摘要

Drilling operations for geothermal and hydrocarbon energy involves technology that controls a highly dynamic and complex process. A transition from assisted control to a higher level of automation not only requires a step-change in technology but also in infrastructure for development and validation of these technologies. The lack of realistic and scalable test environments for automated drilling systems delays qualification of new technology and limits the potential for the industry to reduce costs and minimize the carbon footprint. Since 2016, a high-fidelity drilling simulator has been established and tested for development and validation of Artificial Intelligence (AI) systems for drilling operations. The simulator can be accessed through a web Application Programming Interface (API) and run from a web client or as a Hardware-in-the-loop (HIL) simulator from a control system environment with programmable logic controllers (PLCs). The web enablement makes the simulator suitable for testing AI systems from anywhere in the world without any installation of software. The HIL functionality enables a workflow from early development stages to industrial pilots involving testing in a realistic environment. This paper describes the objectives of the project, the technical solutions, and the results obtained.