By Greg Baird
In the 1998 blockbuster movie, “Armageddon,” when two advanced space shuttles slingshot around the moon at 22,500 miles per hours to land on the global killer asteroid, the pilots see the asteroids debris field and using the FOD (foreign object debris) radar, exclaim “Engaging Evasive Radar!” As the radar scans the horizon, incoming debris is identified, put in a “bounding box” and “auto-coded” in order for the space shuttle to maneuver around the potential collision threat.
NASA tracks more than 500,000 pieces of space debris as they orbit the Earth at 17,500 miles an hour (20,000 pieces are larger than a softball) with still millions of pieces too small to be tracked. Even the international space station resides in a “protected” 30-mile boundary box and maneuvering can take up to 30 hours. Today, we can find a similar technology and process being used to advance sewer collection system asset management.
Computer Vision
This type of human assisted task automation is called computer vision, which is when artificial intelligence (AI) gains a high-level understanding of objects and/or conditions from digital images or videos, such as CCTV data collected from sewer and storm pipeline inspections. Software engineers use machine learning as a way of training an algorithm so that it learns for itself how to fulfill a specific objective, and this happens by exposing the algorithm to various (and numerous) examples that enable it to adjust itself to make improvements over time.
The Sewer Network
A Sanitary Sewer Collection System is a series of pipes, manholes, and lift stations that convey wastewater from homes and businesses to a treatment plant. In the United States, there are more than 800,000 miles of public sewers and 500,000 miles of private laterals. The maintenance goals and objectives for municipal sewer utilities include:
Goal: Maintain sanitary sewer system and related infrastructure to ensure uninterrupted service.
- Objective 1: Annual closed-circuit CCTV inspection of existing sanitary sewer lines and all new collection systems.
- Objective 2: Annually clean the sanitary sewer system.
- Objective 3: Conduct monthly sand/grease interceptor inspections.
- Objective 4: Conduct root cutting for service connections.
Sewer operators have a critical role in managing the public health and the affordability of wastewater services for the community.
Operator assisted sewer AI allows the sewer operator to focus on prioritizing maintenance activities like the type and frequency of cleanings, the timing and location of point repairs, determining the need of additional condition assessment or remaining useful life (RUL) and replacement cost efforts, the evaluation of lining or other rehab trenchless technologies, and the development of the replacement capital plan.
AI Supports Workforce Retention and Training
Wastewater collections operators are able to perform skilled construction, repair and maintenance of wastewater system facilities. They can also operate a wide array of equipment to prevent sewage overflows and blockages. They inspect, clean, maintain, construct and repair wastewater collection systems including sanitary sewers, storm drains, pump stations, pipes, manholes and catch basins, or access points – all of which is aimed at ensuring public health.
AI does not replace the operator or take away the need for certification. AI can accelerate the training for new recruits and leverage the trained operator applying their skills to more value added and meaningful duties. Computer vision software automatically recognizes defects and features in pipes, and assists in generating condition assessment reports at a significantly faster rate (~4x) than if performed manually, and with a higher degree of accuracy (95 to 100 percent) and consistency. This allows the sewer utility, engineers and contractors to save time, money and complete more pipe inspections.
Pipe Condition Assessment Standardization
Standardization provides accuracy, consistency and cost savings when it comes to data management and analytics. While defect coding standardization does not yet exist for water pipes, standardization does exist for sewer pipe condition assessments. The National Association of Sewer Service Companies (NASSCO) trains and certifies technicians and engineers. PACP (Pipeline Assessment Certification Program) is the North American Standard for pipeline defect identification and assessment, providing standardization and consistency to the methods in which pipeline conditions are identified, evaluated and managed. The goal of PACP is to have pipeline system owners create a comprehensive database to properly identify, plan, prioritize, manage, and renovate their pipelines based on sound condition evaluations.
Developed by NASSCO in partnership with the Water Research Center (WRc), PACP, with 226 distinct condition codes, is the North American licensed adaptation of the United Kingdom’s TV inspection coding system developed by the WRc. This coding system is the recognized standard in the United Kingdom and much of Europe and Asia.
NASSCO’s PACP-Based Asset Management
The objective of asset management is aimed at maintaining a function or level of service as cost effectively as possible while maintaining individual components (assets) at the lowest life cycle cost possible. Each sewer asset has a life cycle from design, construction/installation, operations, maintenance, repair/rehab to replacement or disposal. Asset management seeks to continuously identify cost savings and efficiencies in each of the life cycle phases to improve the financial areas of current revenues and expenses for operations and maintenance and capital expenditures typically funded with long term debt.
Good asset management is always integrated with long-term financial planning and rate setting scenarios. PACP asset management seeks answers for the following questions:
- What pipes, manholes and laterals do we own?
- Where are these assets located?
- What are their materials, dimensions, depth and ground cover?
- What is the condition of each asset?
- What other community assets would be affected by failure of a particular asset?
- Which assets are critical to sustained performance?
- What are my best O&M and CIP investment strategies?
- What will be the rehabilitation cost?
- What effect will this have upon the utility budget?
- How should all of this be communicated to stake holders?
PACP can also assist in developing an asset management plan by collecting asset information including:
- Pipe segment length
- Relative location details
- Pipe size
- Pipe shape
- Pipe material
- Upstream manhole data
- Pipe segment lateral data
- Consequence of failure
- Defect codes (Structural and O&M)
- Condition grades
CCTV Direct Assessments are at the Top of the Asset Management Pyramid
Human assisted computer vision using AI/machine learning applied to CCTV data of sewer pipes sits at the top of the sewer network condition assessment and asset management pyramid. These direct condition assessments coded with defect catalogs can quickly, accurately, and effectively provide the basis for all other asset management program decision making.
Good asset management is always integrated with long-term financial planning and rate setting scenarios.
Based on this low-cost AI approach of coding the condition of each sewer pipe, the decision can be made as to where and when additional resources and funding should be allocated for asset work order maintenance and asset management activities moving a utility from a reactive (high cost) mode to a planned and predictive (lower cost) strategic operation. This is especially true for compliance mitigation strategies for consent decrees, the management of sanitary overflows (SSO) and the reduction of the Inflow and Infiltration (I&I) responsible for increased flows to the treatment plant (as high as 45 percent) resulting in higher water treatment costs passed onto rate payers, causing community equity and affordability issues.
Sustainability and Resilience
Sewer condition assessment and asset management also integrates with the sustainability elements of a triple bottom line (TBL) approach evaluating the social, environmental, and both tangible and intangible economic costs. All these efforts build resiliency into the sewer network from risk mitigation to recovery, from climate events or man-made malevolent acts. AI and machine learning will continue to enhance our understanding and operational experiences when it comes to our drinking water, clean water, and reuse infrastructure networks and treatment operations.
Public and private utilities, operators, engineers and contractors each can benefit through the inevitable adoption and acceptance of new technologies like computer vision which better protects our quality of life from the potential threat of collisions from service failures, contamination, and higher costs as our modern day “killer asteroid” called aging water infrastructure continues to challenge our standard of living on Earth.
Greg Baird is president of the Water Finance Research Foundation and a frequent contributor to WF&M. As a management consultant, he specializes in long-term utility planning, infrastructure asset management and capital funding strategies for municipal utilities in the United States. He has served as a municipal finance officer in California and as the CFO of Colorado’s third-largest utility.
