The dent assessment platform that automates the full workflow — from raw ILI data through Level-3 FEA — built on a methodology validated against 79 field excavations with no false negatives in the validation study.
DentSavant automates every step of the dent assessment process — from raw ILI data import through screening, probabilistic strain and fatigue, and Level-3 FEA — with deployment options that support operator security, data-governance, and infrastructure requirements.
Built on an enhanced methodology shaped by systemwide deployment across more than 5,000 dents in a major Canadian liquid transmission system, DentSavant brings operator-grade rigour to pipeline dent programs of any size.
k-Nearest Neighbours machine learning flags anomalous ILI records and achieved 100% recall in the validation dataset, helping prevent incorrect profile data from being used in assessment.
FFT-Gaussian filtering removes ILI measurement noise. 3D profile analysis detects single-apex and multi-apex dent geometry. IMU data integration flags out-of-straightness features that may influence restraint classification and local stress response.
Automatically identifies and evaluates multi-apex dents — a critical gap in many traditional screening tools. High-resolution caliper data often captures complex geometries that single-apex assumptions can misrepresent.
Runs Monte Carlo simulations for each feature to quantify uncertainty in both strain and fatigue assessment. Produces defensible probability-of-failure curves with explicit confidence bounds.
Non-linear FEA with automated 3D shape matching in Abaqus. Converges in 1–3 iterations across diverse dent geometries, achieving RMS shape error of 0.85–2.50 mm. Reduces manual model preparation and accelerates high-throughput Level 3 assessment.
Generates structured assessment reports aligned with API 1183 documentation requirements. Each report includes data provenance, screening rationale, assessment pathway justification, and conclusions structured to support internal review and regulatory documentation.
DentSavant follows a structured, auditable workflow that mirrors the API 1183 decision tree — fully automated, fully traceable.
Import ILI caliper data in standard formats, including CSV and proprietary vendor exports. A k-NN classifier flags records with erroneous channels, geometric outliers, or sensor anomalies and conservatively attempts correction before assessment begins.
FFT-Gaussian filtering removes ILI measurement noise. 3D profile analysis detects single-apex and multi-apex dent geometry. IMU data integration flags out-of-straightness features that may influence restraint classification and local stress response.
Automated API 1183 Level 1 and Level 2 screening is applied across the full dent population. Features that fail screening, or require higher-fidelity assessment, are automatically routed to the appropriate Level 3 FEA workflow.
Monte Carlo simulation quantifies uncertainty in both strain and fatigue assessment. Non-linear FEA with automated mesh and indenter generation runs on local hardware, converging in 1–3 shape-matching iterations per feature.
Structured assessment reports for each dent, ranked priority lists for excavation planning, and population-level summaries. All outputs include full audit trails linking conclusions to input data and methodology citations.
The assessment framework underlying DentSavant was validated against 79 excavated features across a major Canadian liquid transmission system. In that study, the framework showed no false negatives, 85% agreement with excavation findings, and 15% conservative over-calls.
Source: Kainat, M., Virk, A., Yoosef-Ghodsi, N., Bott, S. — "Implementing API 1183 for Dent Assessment," IPC 2022 (IPC2022-87320). Peer-reviewed.
We review your ILI data format, dent population size, operating parameters, and regulatory context to confirm fit and scope.
Run DentSavant on a representative subset of your dent population. Review outputs with our engineers before committing to a full deployment.
License issued for approved machines. Software installed and configured to your data formats and reporting requirements.
Up to 20 hours of virtual training per license year. Annual updates keep your methodology aligned with the latest standards revisions.
Request a technical demo with your own ILI data or a representative sample set. Our engineers walk your team through the workflow, outputs, and deployment approach so you can evaluate fit with confidence.