CCTV Case Study: 99% Increase in Productivity with AI-Assisted Sewer Inspection

The Situation: Aging Infrastructure, Growing Risk

America's wastewater infrastructure is in crisis. The majority of the country's 6 billion feet of sewage pipe is greater than 60 years old, resulting in a host of defects and weaknesses. It's a $1.6 trillion problem that puts communities at risk of environmental spills, collapsing sewer and storm pipes, contaminated drinking water, and more.

Due to consistent lack of funding, wastewater professionals are tasked with proactively managing and maintaining wastewater assets by reviewing CCTV footage — but current methods require human operators to watch every second of a sewer inspection video, which is both costly and prone to error.

• 6 billion feet of sewage pipes across the country
• 60+ year-old infrastructure nearing end of life
• $1.6 trillion problem threatening public health and the environment

The Problem: Wasting the Time of Wastewater Professionals

Current CCTV condition assessment programs use robotic camera equipment to inspect pipe segments one at a time. Effectiveness is limited by human errors of omission, variations in quality and completeness of inspection reports, and subjective interpretations that both over- and underestimate the severity of defects.

Reviewing CCTV footage is a costly drain on time and human resources — and the inconsistency of manual review means asset owners cannot always trust the data they receive.

• 15% underestimation of the severity of sewer defects in manual reviews
• 20% overestimation of the severity of sewer defects in manual reviews

The Solution: Putting the AI in SewerAI

SewerAI developed AutoCode to automatically process inspection videos via deep neural networks and SewerAI's proprietary post-processing algorithms. This AI-driven processing is backed by a manual review of the tool's predictions by NASSCO PACP-certified reviewers — who delete any false positives and enter descriptors, modifiers, clock positions, and percentages associated with certain condition observations into the report.

Head-to-Head Comparison

SewerAI analyzed data from two separate projects completed by a Sewer Services Company (SSC) in Northern California. Both projects used the same make and model of camera equipment and chassis: a CCTV pan & tilt camera mounted on a transporter with 2,000' single-conductor cable, mounted in a 20' Grumman Olson step-van.

• 2019 Project (manual): Cleaning and inspection of 55,462.3 linear feet of 6"–15" Vitrified Clay (VCP) and concrete sanitary sewer pipe.
• 2020 Project (AI-assisted): Cleaning and inspection of 102,270.22 linear feet of 6"–15" VCP and concrete sanitary sewer pipe, located 10 miles from the 2019 project site, with similar access points and pipe configuration.

The Results: Virtually 100% More Efficiency

The productivity comparison between the two projects was striking. By switching from a manual workflow to SewerAI's AI-assisted AutoCode, the SSC nearly doubled its daily inspection output with no additional capital expenditure or software training required.

• 2019 manual workflow: average daily production of 1,352.7 LF of CCTV inspections per crew per shift
• 2020 AI-assisted workflow: average daily production of 2,691.3 LF of CCTV inspections per crew per shift
• 98.95% increase in daily production — virtually doubling field output
• NASSCO-standard deliverables were accepted by asset owners and project stakeholders in both cases
• No additional capital expenditure or training on new software required for the SSC

Additional Benefits of SewerAI's AutoCode

Beyond raw productivity gains, SewerAI's AI-assisted workflow delivers a range of operational and financial benefits for inspection contractors and asset owners:

• Reduces cost per linear foot inspected
• Lowers overall operational costs
• Reduces footage review time significantly
• Doubles field productivity
• Increases review productivity by 70%
• Teams can be onboarded in as little as 1 hour

SewerAI meets NASSCO, PACP, and other industry standards, and is used around the country to serve more than 130 million people.

Download the full CCTV Case Study to see the complete data, methodology, and project details behind this productivity comparison.