PNW Defect Assessment Case Study: Water Utility Finds Maximum Utility With AI
A Pacific Northwest wastewater utility compared SewerAI's AutoCode assessment against its own manual in-house assessments across 29,748 linear feet of CCTV inspections. AI identified 32.99% more conditions and displayed less than 10% margin of error — missing 90.13% fewer conditions than manual survey.
Water Utility Finds Maximum Utility With AI — Saving time, money, and critical infrastructure.
The Problem — When Defects Fly Under the Radar, Underground
A wastewater utility located in the Pacific Northwest, U.S. compared SewerAI's AutoCode-based assessment with that of the utility's in-house, manual assessments. The expectation was that automating tedious, error-prone methods would make the difference between catching disastrous defects or letting them fly under the radar — undetected or misdiagnosed.
- 29,748 linear feet of CCTV inspections
- 132 total surveys conducted
The Solution — Establishing Ground Truth Underground
The assessment completed by AutoCode was also analyzed a second time by SewerAI QC staff to provide a ground truth to compare with the manual assessments. This ensured that the result of the Pipe Survey Comparison Index (PSCI) would accurately reflect inconsistencies in findings between the two methods.
- AutoCode assesses 29,000+ linear feet of CCTV inspections.
- SewerAI QC staff cross-checks the findings against manual assessments.
- A PSCI score is produced to signify inconsistencies between inspection methods.
The Results
- AI identified 32.99% more conditions (defects and features)
- AI displayed <10% margin of error, missing 90.13% fewer conditions than manual survey
Download the full case study here to explore the detailed methodology, data, and outcomes from this engagement.
