Pseudocode and Demonstration of a Multi-Use Artificial Intelligence Algorithm to Perform Challenging and Highly Optimised Pipeline/Cable Routing Cases

Abstract

The process of routing energy conduits (pipelines, cables and umbilicals) in offshore locations represents a critical phase in the concept planning, engineering and construction of these assets. The downstream impact of poorly designed routes is epitomized by a) increased offshore construction durations b) requirements for additional engineered mitigations from geophysical / geotechnical constraints and c) unforeseen requirements for intervention during operations. The cause of these unoptimized routes can be due to low-level engineering tasks which confines to repetitive, inefficient, and unnecessarily iterative processes between draughters, engineers and asset owners. The increasing accessibility and advancement of digital technologies enables highly optimised solutions even through difficult offshore regions. To address the above, this paper presents the scoping, development and application of a multi-functional algorithm created using modern software code frameworks. The algorithm serves as building blocks into an artificial intelligence platform. This routing algorithm simulates, expands and adapts to engineering and consulting expertise from a worldwide network of energy experts. This recreation of expertise firstly identifies commonly encountered routing constraints such as geophysical features, seabed gradients, existing offshore facilities etc. Ideal geometric parameters are then determined to minimise route costs. These processes are then increased, thus enhancing expertise through scale. The algorithm structure will be presented in summarised minimal pseudocode. The pseudocode will present the application programming interface (API) between the constraints based and end parameter calculation approach. The API includes digital innovations such as a) processing of offshore geotechnical survey data, b) recreating offshore locales and routes in a data environment, c) implementation of geospatial intersection detection, d) 3-dimensional route length optimisation and e) automated route selection criteria. This will demonstrate the order of magnitude replication of subject matter expertise into a digital realm, thus eliminating time-consuming, repetition and human error. Finally, the application of the algorithm will be demonstrated by various case studies of offshore locales with challenging conditions such as highly disturbed seabeds and large quantities of existing man-made assets. The front-end cloud platform of the algorithm will be exhibited, showing a streamlined approach and improved routing engineering. Through this, engineers in the future offshore energy developments can answer the question "What is the best route?".