What problem does it solve? Raw as-built pipeline survey files contain null markers, duplicated stations, mixed depth sign conventions, and erroneous points, and the manual spreadsheet cleanup that turns them into usable elevation profiles leaves no record of what was done. This Skill makes that conversion explicit and repeatable, producing a traceable processing log alongside the cleaned profile. ## Core Features & Use Cases - Survey Cleaning and Normalisation: Detects and normalises depth sign conventions, removes sentinel null values and duplicate stations, applies resolution filtering, and flags erroneous points using robust rolling-median residuals. - Span and Cover Screening: Groups consecutive stations into free-span candidates and burial intervals with gap and cover statistics, ready for DNV-RP-F105 or DNV-RP-F114 assessment. - Repeat-Survey Comparison: Interpolates two surveys onto a common KP grid and reports lowering and lifting intervals above a change threshold. - NeqSim Handoff: Resamples the profile into evenly spaced leg positions and elevations for a NeqSim pipe model. - Use Case: An engineer receives a 2010 survey export for a flowline, cleans it, flags a 45 m erroneous point, identifies a free-span candidate near KP 1500, and hands the elevation profile to a NeqSim thermal-hydraulic model. ## Quick Start Ask the agent to process your survey records with PipelineSurveyProcessor using a minimum KP spacing and span gap threshold, then print the flagged points, span candidates, and processing log.